Research Network for
Self-Organization of Robot Behavior
Literature

Inhalt

Bibfiles

Literaturliste (fremde Paper)

[1] J. Kingdon. Intelligent Systems and Financial Forecasting. Springer, 11997.
[ bib ]
[2] N. Hamed. Self-Referential Dynamical Systems and Developmental Robotics. PhD thesis, University of Leipzig, 2006. In preparation.
[ bib ]
[3] L. Righetti and I. A.J. Programmable central pattern generators: an application to biped locomotion control. In Proceedings of the 2006 IEEE International Conference on Robotics and Automation, 2006. In press.
[ bib | www ]
[4] J. Buchli, L. Righetti, and A. Ijspeert. A dynamical systems approach to learning: a frequency-adaptive hopper robot. In Proceedings of the VIIIth European Conference on Artificial Life ECAL 2005, Lecture Notes in Artificial Intelligence, pages 210-220. Springer Verlag, 2005.
[ bib | www ]
[5] S. H. Collins, A. Ruina, R. Tedrake, and M. Wisse. Efficient bipedal robots based on passive-dynamic walkers. Science, 307:1082-1085, 2005.
[ bib ]
[6] I. Harvey, E. Di Paolo, R. Wood, M. Quinn, and E. A. Tuci. Evolutionary robotics: A new scientific tool for studying cognition. Artificial Life, 11(1-2):79-98, 2005.
[ bib ]
[7] R. Haschke and J. J. Steil. Input space bifurcation manifolds of recurrent neural networks. Neurocomputing, 64C:25-38, 2005.
[ bib ]
[8] F. Iida, G. J. Gomez, and R. Pfeifer. Exploiting body dynamics for controlling a running quadruped robot. In International Conference of Advanced Robotics (ICAR 05), 2005.
[ bib ]
[9] M. Komosinski. Computer science in evolutionary biology and robotics: Framsticks artificial life. 2005.
[ bib ]
[10] R. Legenstein and W. Maass. What makes a dynamical system computationally powerful? In S. Haykin, J. C. Principe, T. Sejnowski, and J. McWhirter, editors, New Directions in Statistical Signal Processing: From Systems to Brain. MIT Press, 2005. to appear.
[ bib ]
[11] R. A. Legenstein and W. Maass. Edge of chaos and prediction of computational power for neural microcircuit models. submitted for publication, 2005.
[ bib ]

What makes a neural microcircuit computationally powerful? Or more precisely, which measurable quantities could explain why one microcircuit C is better suited for a particular family of computational tasks than another microcircuit C'? One potential answer comes from results on cellular automata and random Boolean networks, where some evidence was provided that their computational power for offline computations is largest at the edge of chaos, i.e. at the transition boundary between order and chaos. We analyse in this article the significance of the edge of chaos for real time computations in neural microcircuit models consisting of spiking neurons and dynamic synapses. We find that the edge of chaos predicts quite well those values of circuit parameters that yield maximal computational power. But obviously it makes no prediction of their computational power for other parameter values. Therefore, we propose a new method for predicting the computational power of neural microcircuit models. The new measure estimates directly the kernel property and the generalization capability of a neural microcircuit. We validate the proposed measure by comparing its prediction with direct evaluations of the computational performance of various neural microcircuit models. This procedure is applied first to microcircuit models that differ with regard to the spatial range of synaptic connections and their strength, and then to microcircuit models that differ with regard to the level of background input currents, the conductance, and the level of noise on the membrane potential of neurons. In the latter case the proposed method allows us to quantify differences in the computational power and generalization capability of neural circuits in different dynamic regimes (UP- and DOWN-states) that have been demonstrated through intracellular recordings in vivo.
[12] R. Liebscher. Echtzeitfähige Lernverfahren Für Kooperierende Roboter. PhD thesis, University Leipzig, Institute of Computer Science, 2005.
[ bib ]
[13] G. Martius. Aktives lernen in der sensomotorischen schleife. Master's thesis, Universität Leipzig, 2005.
[ bib ]
[14] R. Smith. Open dynamics engine. http://ode.org/, 2005.
[ bib ]
[15] P.-Y. Oudeyer. The self-organization of speech sounds. Journal of Theoretical Biology, 233(3):435-449, 2005.
[ bib ]
[16] A. Cangelosi. The sensorimotor bases of linguistic structure: Experiments with grounded adaptive agents. In S. S. et al., editor, SAB04, pages 487-496, Los Angeles, July 2004. Cambridge MA, MIT Press. Proceedings of the Eighth International Conference on the Simulation of Adaptive Behaviour: From Animals to Animats 8.
[ bib ]
[17] M. Rosencrantz, G. Gordon, and S. Thrun. Learning low dimensional predictive representations. In ICML '04: Twenty-first international conference on Machine learning. ACM Press, 2004.
[ bib ]
[18] A. G. Brooks, J. Gray, G. Hoffman, A. Lockerd, H. Lee, and C. Breazeal. Robot's play: interactive games with sociable machines. Comput. Entertain., 2(3):10-10, 2004.
[ bib ]
[19] B. Bartlett, V. Estivill-Castro, and S. Seymon. Dogs or robots: why do children see them as robotic pets rather than canine machines? In CRPIT '28: Proceedings of the fifth conference on Australasian user interface, pages 7-14. Australian Computer Society, Inc., 2004.
[ bib ]
[20] J. Hoffmann and U. Düffert. Frequency space representation and transitions of quadruped robot gaits. In CRPIT '26: Proceedings of the 27th conference on Australasian computer science, pages 275-278. Australian Computer Society, Inc., 2004.
[ bib ]
[21] L. Berthouze and M. Lungarella. Robot Bouncing: On the Synergy Between Neural and Body-Environment Dynamics, volume 3139 of Lecture Notes in Computer Science, pages 86-97. Springer Verlag, embodied artificial intelligence edition, 2004.
[ bib ]
[22] T. Buehrmann and E. Di Paolo. Closing the loop: evolving a model-free visually guided robot arm. In Proceedings of the ninth international conference on the simulation and synthesis of living systems, volume 9. MIT Press, 2004.
[ bib ]
[23] M. Bunk. Echtzeitfähige modellbasierte bilderkennung zur visuomotorischen kontrolle bewegter objekte. Master's thesis, Universität Leipzig, 2004.
[ bib ]
[24] G. Deutschmann. Echtzeitfähige lernverfahren zur situationserkennung aus komplexen sensordaten für autonome roboter. Technical report, Universität Leipzig, 2004. Diplomarbeit.
[ bib ]
[25] D. Kim. Self-organization for multi-agent groups. International Journal of Control, Automation, and Systems, 2:333-342, 2004.
[ bib ]
[26] K. Lerman, A. Martinoli, and A. Galstyan. A review of probabilistic macroscopic models for swarm robotic systems. Self-organization of Adaptive Behavior04, Santa Monica, CA, 2004.
[ bib ]
[27] K. Lerman and A. Galstyan. Automatically modeling group behavior of simple agents. Agent Modeling Workshop, AAMAS-04, New York, NY., 2004.
[ bib ]
[28] B. Libet. Mind Time : The Temporal Factor in Consciousness. Harvard University Press, 2004.
[ bib ]
[29] B. Lindner, J. Garcia-Ojalvo, A. Neiman, and L. Schimansky-Geier. Effects of noise in excitable systems. Phys. Report, 392,:321-424, 2004.
[ bib ]
[30] M. Lungarella, G. Metta, R. Pfeifer, and G. Sandini. Developmental robotics: a survey. Connection Science, 0(0):1-40, 2004.
[ bib | .pdf ]

Developmental robotics is an emerging field located at the intersection of robotics, cognitive science and developmental sciences. This paper elucidates the main reasons and key motivations behind the convergence of fields with seemingly disparate interests, and shows why developmental robotics might prove to be beneficial for all fields involved. The methodology advocated is synthetic and two-pronged: on the one hand, it employs robots to instantiate models originating from developmental sciences; on the other hand, it aims to develop better robotic systems by exploiting insights gained from studies on ontogenetic development. This paper gives a survey of the relevant research issues and points to some future research directions.

Keywords: developmental robotics, artificial intelligence
[31] I. Macinnes and E. Di Paolo. Crawling out of the simulation: Evolving real robot morphologies using cheap, reusable modules. In J. Pollack, M. Bedau, P. Husbands, T. Ikegami, and R. Watson, editors, Proceedings of the Ninth International Conference on the Simulation and Synthesis of Living Systems, ALIFE'9. MIT Press, Cambridge MA, 2004.
[ bib ]
[32] J. Nakanishi, J. Morimoto, G. Endo, G. Cheng, S. Schaal, and M. Kawato. Learning from demonstration and adaptation of biped locomotion. Robotics and Autonomous Systems, 47:79-91, 2004.
[ bib ]
[33] J. Nakanishi and S. Farrell, J. A.and Schaal. Learning composite adaptive control for a class of nonlinear systems. IEEE International Conference on Robotics and Automation, pages 2647-2652, 2004.
[ bib ]
[34] C. Maurer, T. Mergner, and R. J. Peterka. Abnormal resonance behavior of the postural control loop in parkinson's disease. Exp Brain Res., 157(3):369-76, 2004.
[ bib ]
[35] F. Iida, R. Pfeifer, L. Steels, and Y. Kuniyoshi, editors. Embodied Artificial Intelligence, International Seminar, Dagstuhl Castle, Germany, July 7-11, 2003, Revised Papers, volume 3139 of Lecture Notes in Computer Science. Springer, 2004.
[ bib ]
[36] J. Popp. Sphericalrobots. http://www.sphericalrobots.com, 2004.
[ bib ]
[37] R. Siegwart and I. R. Nourbakhsh. Autonomous Mobile Robots. MIT, 2004.
[ bib ]
[38] J. Tani. Symbols and dynamics in embodied cognition: Revisiting a robot experiment. In M. V. Butz, O. Sigaud, and P. Gerard, editors, Anticipatory Behavior in Adaptive Learning Systems, pages 167-178. Springer-Verlag, 2004.
[ bib ]
[39] R. Tedrake, T. W. Zhang, and H. S. Seung. Stochastic policy gradient reinforcement learning on a simple 3d biped. In Proceedings of the IEEE International Conference on Intelligent Robots and Systems (IROS), pages 2849-2854, 2004.
[ bib ]
[40] G. G. Turrigiano and S. B. Nelson. Homeostatic plasticity in the developing nervous system. Nature Reviews Neuroscience, 5:97-107, 2004.
[ bib ]
[41] H. Williams. Homeostatic plasticity in recurrent neural networks. In S. Schaal, editor, From Animals to Animats: Proceedings of the 8th Intl. Conf. On Simulation of Adaptive Behavior, volume 8 of 8. MIT Press, 2004.
[ bib ]
[42] R. Haschke. Bifurcations in Discrete-Time Neural Networks - Controlling Complex Network Behaviour with Inputs. PhD thesis, Bielefeld University, Sep 2003.
[ bib ]
[43] M. Rosencrantz, G. Gordon, and S. Thrun. Locating moving entities in indoor environments with teams of mobile robots. In AAMAS '03: Proceedings of the second international joint conference on Autonomous agents and multiagent systems, pages 233-240. ACM Press, 2003.
[ bib ]
[44] M. N. Nicolescu and M. J. Mataric. Natural methods for robot task learning: instructive demonstrations, generalization and practice. In AAMAS '03: Proceedings of the second international joint conference on Autonomous agents and multiagent systems, pages 241-248. ACM Press, 2003.
[ bib ]
[45] H. Moravec. Robots, after all. Commun. ACM, 46(10):90-97, 2003.
[ bib ]
[46] S. Pierre, M. Barbeau, and E. Kranakis, editors. Ad-Hoc, Mobile, and Wireless Networks, Second International Conference, ADHOC-NOW 2003 Montreal, Canada, October 8-10, 2003, Proceedings, volume 2865 of Lecture Notes in Computer Science. Springer, 2003.
[ bib ]
[47] A. Bastian, G. Schöner, and A. Riehle. Preshaping and continuous evolution of motor cortical representations during movement preparation. European Journal of Neuroscience, 18:2047-2058, 2003.
[ bib | .pdf ]
[48] J. Burrone and V. N. Murthy. Synaptic gain control and homeostasis. Current Opinion in Neurobiology, 13:560-567, 2003.
[ bib ]
[49] H. Cruse. The evolution of cognition - a hypothesis. . Cognitive Science, 27(1):135-155, 2003.
[ bib ]
[50] D. J. Bakkum, A. C. Shkolnik, G. Ben-Ary, P. Gamblen, T. B. DeMarse, and S. M. Potter. Removing some 'a' from ai: Embodied cultured networks. In Pierre et al. [46], pages 130-145.
[ bib ]
[51] V. V. Hafner. Agent-environment interaction in visual homing. In Pierre et al. [46], pages 180-185.
[ bib ]
[52] O. Holland. The future of embodied artificial intelligence: Machine consciousness?. In Pierre et al. [46], pages 37-53.
[ bib ]
[53] K. Hosoda. Robot finger design for developmental tactile interaction: Anthropomorphic robotic soft fingertip with randomly distributed receptors. In Pierre et al. [46], pages 219-230.
[ bib ]
[54] F. Iida and R. Pfeifer. Self-stabilization and behavioral diversity of embodied adaptive locomotion. In Pierre et al. [46], pages 119-129.
[ bib ]
[55] A. Ishiguro and T. Kawakatsu. How should control and body systems be coupled? a robotic case study. In Pierre et al. [46], pages 107-118.
[ bib ]
[56] L. Lichtensteiger. The need to adapt and its implications for embodiment. In Pierre et al. [46], pages 98-106.
[ bib ]
[57] M. Lungarella and L. Berthouze. Robot bouncing: On the synergy between neural and body-environment dynamics. In Pierre et al. [46], pages 86-97.
[ bib ]
[58] S. Murata, A. Kamimura, H. Kurokawa, E. Yoshida, K. Tomita, and S. Kokaji. Self-reconfigurable robots: Platforms for emerging functionality. In Pierre et al. [46], pages 312-330.
[ bib ]
[59] R. Pfeifer and F. Iida. Embodied artificial intelligence: Trends and challenges. In Pierre et al. [46], pages 1-26.
[ bib ]
[60] O. Sporns and T. K. Pegors. Information-theoretical aspects of embodied artificial intelligence. In Pierre et al. [46], pages 74-85.
[ bib ]
[61] L. Steels. The autotelic principle. In Pierre et al. [46], pages 231-242.
[ bib ]
[62] T. Ziemke. Embodied ai as science: Models of embodied cognition, embodied models of cognition, or both? In Pierre et al. [46], pages 27-36.
[ bib ]
[63] E. D. Paolo. Organismically-inspired robotics: Homeostatic adaptation and natural teleology beyond the closed sensorimotor loop. In K. Murase and T. Asakura, editors, Dynamical Systems Approach to Embodiment and Sociality, pages 19 - 42, Adelaide, 2003. Advanced Knowledge International.
[ bib ]
[64] E. Di Paolo. Evolving spike-timing-dependent plasticity for single-trial learning in robots. Philosophical Transactions of the Royal Society of London, Series A: Mathematical, Physical and Engineering Sciences, 361(1811):2299 - 2319, 2003.
[ bib ]
[65] B. R. Fajen, W. H. Warren, S. Temizer, and L. P. Kaelbling. A dynamical model of visually-guided steering, obstacle avoidance, and route selection. Int. J. Comput. Vision, 54(1-3):13-34, 2003.
[ bib ]
[66] Y. Fukoka, H. Kimura, and A. Cohen. Adaptive dynamic walking of a quadruped robot on irregular terrain based on biological concepts. International Journal of Robotics Research, 22(3-4):187-202, 2003.
[ bib ]
[67] H. S. Hock, G. Schöner, and M. A. Giese. The dynamical foundations of motion pattern formation: Stability, selective adaptation, and perceptual continuity. Perception & Psychophysics, 65:429-457, 2003.
[ bib ]
[68] B. Porr, C. von Ferber, and F. Wörgötter. ISO learning approximates a solution to the inverse-controller problem in an unsupervised behavioral paradigm. Neural Computation, 15(4):865-884, 2003.
[ bib ]
[69] B. Porr and F. Wörgötter. Isotropic sequence order learning. Neural Computation, 15(4):831-864, 2003.
[ bib | http ]
[70] F. Kaplan and P.-Y. Oudeyer. Maximizing learning progress: An internal reward system for development. In Pierre et al. [46], pages 259-270.
[ bib ]
[71] D.-H. Kim and J.-H. Kim. A real-time limit-cycle navigation method for fast mobile robots and its application to robot soccer. Robot. Auton. Syst., 42(1):17-30, 2003.
[ bib ]

Summary: A mobile robot should be designed to navigate with collision avoidance capability in the real world, flexibly coping with the changing environment. In this paper, a novel limit-cycle navigation method is proposed for a fast mobile robot using the limit-cycle characteristics of a 2nd-order nonlinear function. It can be applied to the robot operating in a dynamically changing environment, such as in a robot soccer system. By adjusting the radius of the motion circle and the direction of obstacle avoidance, the navigation method proposed enables a robot to maneuver smoothly towards any desired destination. Simulations and real experiments using a robot soccer system demonstrate the merits and practical applicability of the proposed method.

Keywords: robot soccer; navigation; limit cycle; real-time control
[72] Y. Kuniyoshi, Y. Yorozu, Y. Ohmura, K. Terada, T. Otani, A. Nagakubo, and T. Yamamoto. From humanoid embodiment to theory of mind. In Pierre et al. [46], pages 202-218.
[ bib ]
[73] S. Mitchell and R. A. Silver. Shunting inhibition modulates neuronal gain during synaptic excitation. Neuron, 38:433-45., 2003.
[ bib ]
[74] U. Nehmzow and K. Walker. Quantitative description of robot-environment interaction using chaos theory. In Proc. European Conference on Mobile Robotics (ECMR), 2003.
[ bib ]
[75] F. Pasemann, M. Hild, and K. Zahedi. SO(2)-networks as neural oscillators. In J. Mira and J. Alvarez, editors, Computational Methods in Neural Modeling, pages 144-151, Berlin, Heidelberg, New York, 2003. Springer.
[ bib ]
[76] R. Pfeifer and F. Iida. Embodied artificial intelligence: Trends and challenges. In Embodied Artificial Intelligence, pages 1-26, Cambridge, MA, 2003. Bradford Books, MIT Press.
[ bib ]
[77] D. Philipona, J. O'Regan, and J. Nadal. Is there something out there? inferring space from sensorimotor dependencies. Neural Computation, 15(9):2029-49, 2003.
[ bib ]
[78] S. Salenius and R. Hari. Synchronous cortical oscillatory activity during motor control. Current Opinion in Neurobiology, 13:678, 2003.
[ bib ]
[79] J. Peters, S. Vijayakumar, and S. Schaal. Reinforcement learning for humanoid robotics. In Humanoids2003, Third IEEE-RAS International Conference on Humanoid Robots., 2003.
[ bib ]
[80] L. Steels. Evolving grounded communication for robots. Trends in Cognitive Sciences, 7, 2003.
[ bib ]
[81] J. Tani and M. Ito. Self-organization of behavioral primitives as multiple attractor dynamics: A robot experiment. IEEE Transactions of on Systems, Man, and Cybernetics Part A: Systems and Humans, 33(4):481-488, 2003.
[ bib ]
[82] W. Tschacher and J. Dauwalder. The Dynamical Systems Approach to Cognition: Concepts and Empirical Paradigms Based on Self-Organization, Embodiment, and Coordination Dynamics. World Scientific Publishing Company, Singapore, 2003.
[ bib ]
[83] H. Yokoi, A. H. Arieta, R. Katoh, W. Yu, I. Watanabe, and M. Maruishi. Mutual adaptation in a prosthetics application. In Pierre et al. [46], pages 146-159.
[ bib ]
[84] H. Bowman, A. Aron, F. Schlaghecken, and M. Eimer. Neural network modelling of inhibition in visuo-motor control. In J. A. B. . W. Lowe, editor, Proceedings of the Seventh Neural Computation and Psychology Workshop: Connectionist Models of Cognition and Perception, pages 209-222. World Scientific, September 2002.
[ bib | http ]
[85] S. Thrun. Probabilistic robotics. Commun. ACM, 45(3):52-57, 2002.
[ bib ]
[86] K. H. Low, W. K. Leow, and J. Marcelo H. Ang. A hybrid mobile robot architecture with integrated planning and control. In AAMAS '02: Proceedings of the first international joint conference on Autonomous agents and multiagent systems, pages 219-226. ACM Press, 2002.
[ bib ]
[87] P. Andry, P. Gaussier, and J. Nadel. From sensorimotor coordination to low level imitation. In Second international workshop on epigenetic robotics, EPIROB 02, pages 7-15, 2002.
[ bib | http ]
[88] N. Berglund and B. Gentz. Pathwise description of dynamic pitchfork bifurcations with additive noise. Probab. Theory Related Fields, 122:341-388, 2002.
[ bib ]
[89] B. Porr and F. Wörgötter. Learning a forward model of a reflex. In NIPS, pages 1531-1538, 2002.
[ bib ]
[90] E. Di Paolo. Spike-timing dependent plasticity for evolved robots. Adaptive Behavior, 10(3/4):243-263, 2002.
[ bib ]
[91] K. Doya, K. Samejima, K. Katagiri, and K. M. Kawato. Multiple model based reinforcement learning. Neural Computation, 14:1347-1369, 2002.
[ bib ]
[92] D. E. Feldman. Synapses, scaling and homeostasis in vivo. Nature Neuroscience, 5:712 - 714, 2002.
[ bib ]
[93] W. Gerstner and W. M. Kistler. Spiking Neuron Models: Single Neurons, Populations, Plasticity. Cambridge University Press, New York, 2002.
[ bib ]
[94] M. Hülse and F. Pasemann. Dynamical Neural Schmitt Trigger for Robot Control, volume 2415 of Lecture Notes in Computer Science. Springer, Berlin, Heidelberg, New York, 2002.
[ bib ]
[95] S. A. Kauffman. Investigations. Oxford University Press, 2002.
[ bib ]
[96] K. Labusch and D. Polani. Sensor evolution in a homeokinetic system. In Polani et al. [104], pages 199-208.
[ bib ]
[97] W. Maass and H. Markram. Synapses as dynamic memory buffers. Neural Netw., 15(2):155-161, 2002.
[ bib ]
[98] L. Moreau and E. Sontag. Balancing at the border of instability. submitted, 2002.
[ bib ]
[99] L. Moreau, E. Sontag, and M. Arcak. Feedback tuning of bifurcations. submitted, 2002.
[ bib ]
[100] F. Pasemann. Complex dynamics and the structure of small neural networks network. Computation in Neural Systems, 13:195-216, 2002.
[ bib ]
[101] M. Hülse and F. Pasemann. Dynamical Neural Schmitt Trigger for Robot Control, volume 2415 of Lecture Notes in Computer Science. Springer, 2002.
[ bib ]
[102] P. Perruchet and A. Vinter. The self-organizing consciousness. Behav Brain Sci., 25:297-330, 2002.
[ bib ]
[103] D. Polani, J. Kim, and T. Martinetz, editors. Fifth German Workshop on Artificial Life, March 18-20, 2002, Lübeck, Germany. IOS Press Infix, Aka, 2002.
[ bib ]
[104] D. Polani, J. Kim, and T. Martinetz, editors. Fifth German Workshop on Artificial Life, March 18-20, 2002, Lübeck, Germany. IOS Press Infix, Aka, 2002.
[ bib ]
[105] L. Steels. Grounding symbols through evolutionary language games. In A. Cangelosi and D. Parisi, editors, Simulating the Evolution of Language, chapter 10, pages 211-226. Springer Verlag, London, 2002.
[ bib | .html ]
[106] L. Steels and F. Kaplan. Bootstrapping grounded word semantics. In T. Briscoe, editor, Linguistic evolution through language acquisition: Formal and Computational Models, pages ??-?? Cambridge University Press, Cambridge, UK, 2002.
[ bib | .ps ]
[107] L. Steels. Grounding symbols through evolutionary language games. In A. Cangelosi and D. Parisi, editors, Simulating the Evolution of Language, pages 211-226. Springer-Verlag, Berlin, 2002.
[ bib ]
[108] L. Steels. Language games for emergent semantics. IEEE Intelligent Systems, 17(1):83-85, 2002.
[ bib ]
[109] L. Wiskott and T. Sejnowski. Slow feature analysis: Unsupervised learning of invariances. Neural Computation, 14(4):715-770, 2002.
[ bib ]

Invariant features of temporally varying signals are useful for analysis and classification. Slow feature analysis (SFA) is a new method for learning invariant or slowly varying features from a vectorial input signal. SFA is based on a non-linear expansion of the input signal and application of principal component analysis to this expanded signal and its time derivative. It is guaranteed to find the optimal solution within a family of functions directly and can learn to extract a large number of decorrelated features, which are ordered by their degree of invariance. SFA can be applied hierarchically to process high dimensional input signals and to extract complex features. Slow feature analysis is applied first to complex cell tuning properties based on simple cell output including disparity and motion. Then, more complicated input-output functions are learned by repeated application of SFA. Finally, a hierarchical network of SFA-modules is presented as a simple model of the visual system. The same unstructured network can learn translation, size, rotation, contrast, or, to a lesser degree, illumination invariance for one-dimensional objects, depending only on the training stimulus. Surprisingly, only a few training objects sufficed to achieve good generalization to new objects. The generated representation is suitable for object recognition. Performance degrades, if the network is trained to learn multiple invariances simultaneously.
[110] T. Shibata and S. Schaal. Biomimetic gaze stabilization based on feedback-error-learning with nonparametric regression networks. Neural Netw., 14(2), 2001.
[ bib ]
[111] M. Yoshizaki, Y. Kuno, and A. Nakamura. Human-robot interface based on the mutual assistance between speech and vision. In PUI '01: Proceedings of the 2001 workshop on Percetive user interfaces, pages 1-4. ACM Press, 2001.
[ bib ]
[112] A. Balaam. nbrains: A new type of robot brain. In J. Kelemen and P. Sosek, editors, Advances in Artificial Life : 6th European Conference, ECAL 2001 , Prague, Czech Republic, September 10-14, 2001, Proceedings, page 491. Springer-Verlag Heidelberg, 2001.
[ bib ]
[113] W. Bialek, I. Nemenman, and N. Tishby. Predictability, complexity and learning. Neural Computation, 13:2409, 2001.
[ bib | www ]
[114] G. Bi and M. M. Poo. Synaptic modification by correlated activity: Hebb's postulate revisited. Annu Rev Neurosci, 24:139-66, 2001.
[ bib ]
[115] A. Bredenfeld, H. Jaeger, and T. Christaller. Mobile robots with dual dynamics. ERCIM News, 42, 2001.
[ bib ]
[116] S. Collins, M. Wisse, and A. Ruina. A 3-d passive dynamic walking robot with two legs and knees. International Journal of Robotics Research, 20 (7):607-615, 2001.
[ bib ]
[117] B. Porr and F. Wörgötter. Temporal hebbian learning in rate-coded neural networks: A theoretical approach towards classical conditioning. In ICANN, pages 1115-1120, 2001.
[ bib | www ]
[118] R. Haschke, J. J. Steil, and H. Ritter. Controlling Oscillatory Behaviour of a Two Neuron Recurrent Neural Network Using Inputs, volume 2130 of Lecture Notes in Computer Science. Springer, 2001.
[ bib ]
[119] R. Haschke, J. J. Steil, and H. Ritter. Controlling oscillatory behavior of a two neuron recurrent neural network using inputs. In Proc. of the Int. Conf. on Artificial Neural Networks (ICANN), Wien, Austria, 2001.
[ bib ]
[120] L. Steels. Language games for autonomous robots. IEEE Intelligent Systems, 16(5):16-22, 2001.
[ bib | .ps ]
[121] J. Tani. Learning to generate articulated behavior through the bottom-up and the top-down interaction processes. Neural Networks. The Official Journal of the International Neural Network Society, European Neural Network Society, Japanese Neural Network Society., 16(1):11-23, 2001.
[ bib ]

Summary: A novel hierarchical neural network architecture for sensory-motor learning and behavior generation is proposed. Two levels of forward model neural networks are operated on different time scales while parametric interactions are allowed between the two network levels in the bottom-up and top-down directions. The models are examined through experiments of behavior learning and generation using a real robot arm equipped with a vision system. The results of the learning experiments showed that the behavioral patterns are learned by self-organizing the behavioral primitives in the lower level and combining the primitives sequentially in the higher level. The results contrast with prior work by Pawelzik et al. [Neural Comput. 8, 340 (1996)], Tani and Nolfi [From animals to animats, 1998], and Wolpert and Kawato [Neural Networks 11, 1317 (1998)] in that the primitives are represented in a distributed manner in the network in the present scheme whereas, in the prior work, the primitives were localized in specific modules in the network. Further experiments of on-line planning showed that the behavior could be generated robustly against a background of real world noise while the behavior plans could be modified flexibly in response to changes in the environment. It is concluded that the interaction between the bottom-up process of recalling the past and the top-down process of predicting the future enables both robust and flexible situated behavior.

Keywords: Learning; Behavior; Articulation; Chunking
[122] A. Billard and M. J. Mataric. A biologically inspired robotic model for learning by imitation. In AGENTS '00: Proceedings of the fourth international conference on Autonomous agents, pages 373-380. ACM Press, 2000.
[ bib ]
[123] R. D. Beer. Dynamical approaches to cognitive science. Trends in Cognitive Sciences, 4(3):91-99., 2000.
[ bib ]
[124] G. Chechik and N. Tishby. Temporally dependent plasticity: An information theoretic account. In NIPS, pages 110-116, 2000.
[ bib ]
[125] M. Komosinski. The world of framsticks: Simulation, evolution, interaction. In Proceedings of 2nd International Conference on Virtual Worlds (VW2000), Paris, Lecture Notes in Artificial Intelligence 1834, pages 214-224., Berlin, Heidelberg, New York, 2000. Springer-Verlag.
[ bib ]
[126] S. Kotosaka and S. Schaal. Synchronized robot drumming by neural oscillators. In Proceedings of the International Symposium on Adaptive Motion of Animals and Machines, 2000.
[ bib ]
[127] W. M. Kistler and J. L. van Hemmen. Modeling synaptic plasticity in conjuction with the timing of pre- and postsynaptic action potentials. Neural Computation, 12(2):385-405, 2000.
[ bib ]
[128] S. Nolfi and D. Floreano. Evolutionary Robotics. MIT Press,Cambridge, MA, 2000.
[ bib ]
[129] E. D. Paolo. Homeostatic adaptation to inversion in the visual field and other sensorimotor disruptions, 2000.
[ bib ]
[130] R. J. Peterka. Postural control model interpretation of stabilogram diffusion analysis. Biol Cybern, 82:335-34382, 2000.
[ bib ]
[131] S. Schaal and S. Kotosaka. Synchronized robot drumming by neural oscillators. In The International Symposium on Adaptive Motion of Animals and Machines, Montreal, 2000.
[ bib ]
[132] S. Song, K. Miller, and L. Abbott. Competitive hebbian learning through spiketime -dependent synaptic plasticity. Nat. Neurosci., 3:919-926, 2000.
[ bib ]
[133] S. Song, K. Miller, and L. Abbott. Competitive hebbian learning through spiketime -dependent synaptic plasticity. Nat. Neurosci., 3:919-926, 2000.
[ bib ]
[134] R. Sutton, D. McAllester, S. Singh, and Y. Mansour. Policy gradient methods for reinforcement learning with function approximation. Advances in Neural Information Processing Systems, 12:1057-1063, 2000.
[ bib ]
[135] G. Taga. Nonlinear dynamics of the human motor control - real-time and anticipatory adaptation of locomotion and development of movements. In Proceedings of the International Symposium on Adaptive Motion of Animals and Machines, 2000.
[ bib ]
[136] G. G. Turrigiano and S. Nelson. Hebb and homeostasis in neuronal plasticity. Current Opinion in Neurobiology, 10:358-364, 2000.
[ bib ]

The positive-feedback nature of Hebbian plasticity can destabilize the properties of neuronal networks. Recent work has demonstrated that this destabilizing influence is counteracted by a number of homeostatic plasticity mechanisms that stabilize neuronal activity. Such mechanisms include global changes in synaptic strengths, changes in neuronal excitability, and the regulation of synapse number. These recent studies suggest that Hebbian and homeostatic plasticity often target the same molecular substrates, and have opposing effects on synaptic or neuronal properties. These advances significantly broaden our framework for understanding the effects of activity on synaptic function and neuronal excitability.
[137] R. Brooks. Cambrian intelligence: The early history of the new ai, 1999.
[ bib ]
[138] S. Nolfi and J. Tani. Extracting regularities in space and time through a cascade of prediction networks: The case of a mobile robot navigating in a structured environment. Connection Science, Journal of Neural Computing, Artificial Intelligence and Cognitive Research, 11(2):125-148, 1999.
[ bib ]

Summary: We propose that the ability to extract regularities from time series through prediction learning can be enhanced if we use a hierarchical architecture in which higher layers are trained to predict the internal state of lower layers when such states change significantly. This hierarchical organization has two functions: (a) it forces the system to recode sensory information progressively so as to enhance useful regularities and filter out useless information; and (b) it progressively reduces the length of the sequences which should be predicted going from lower to higher layers. This, in turn, allows higher levels to extract higher-level regularities which are hidden at the sensory level. By training an architecture of this type to predict the next sensory state of a robot navigating in an environment divided in two rooms, we show how the first-level prediction layer extracts low-level regularities, while the second-level prediction layer extracts higher-level regularities.

Keywords: prediction learning; development; mobile robotics
[139] G. Deboeck and T. Kohonen. Visual Explorations in Finance with Self-Organizing Maps. Springer, 1999.
[ bib ]
[140] K. Doya. Reinforcement learning in continuous time and space. Neural Computation, 12:243-269, 1999.
[ bib ]
[141] S. Viollet and N. Franceschini. Visual servo system based on a biologically-inspired scanning sensor. In Sensor Fusion and decentralized Control in Robotic Systems II, volume SPIE Vol. 3839, pages 144-155, Bellingham, USA, 1999.
[ bib ]
[142] K. Hase and N. Yamazaki. Computational evolution of human bipedal walking by a neuro-musculo-skeletal model. Artificial Life and Robotics, 3:133-138, 1999.
[ bib ]
[143] B. Libet, A. Freeman, and J. K. B. Sutherland. The Volitional Brain - Towards a Neuroscience of Free Will. Imprint Academic, 1999.
[ bib ]
[144] W. Maass. Computing with spiking neurons. In W. Maass and C. M. Bishop, editors, Pulsed Neural Networks, pages 55-85. MIT Press (Cambridge), 1999.
[ bib ]
[145] W. M. Kistler and J. L. van Hemmen. Short-term synaptic plasticity and network behavior. Neural Computation, 11(7):1579-1594, 1999.
[ bib ]
[146] R. Pfeifer and C. Scheier. Understanding Intelligence. Bradford Books, 1999.
[ bib ]
[147] R. Pfeiffer and C. Scheier. Understanding Intelligence. MIT Press,Cambridge, MA, 1999.
[ bib ]
[148] L. Steels. How language bootstraps cognition. In I. Wachsmutt and B. Jung, editors, KogWis99. Proceedings der 4. Fachtagung der Gesellschaft für Kognitionswissenschaft, pages 1-3, Braunschweig, Germany, 1999. Infix.
[ bib ]
[149] G. G. Turrigiano. Homeostatic plasticity in neuronal networks: The more things change, the more they stay the same. Trends in Neuroscience, 22:221-228, 1999.
[ bib ]

During learning and development, neural circuitry is refined, in part, through changes in the number and strength of synapses. Most studies of long-term changes in synaptic strength have concentrated on Hebbian mechanisms, where these changes occur in a synapse-specific manner. While Hebbian mechanisms are important for modifying neuronal circuitry selectively, they might not be sufficient because they tend to destabilize the activity of neuronal networks. Recently, several forms of homeostatic plasticity that stabilize the properties of neural circuits have been identified. These include mechanisms that regulate neuronal excitability, stabilize total synaptic strength, and influence the rate and extent of synapse formation. These forms of homeostatic plasticity are likely to go 'hand-in-glove' with Hebbian mechanisms to allow experience to modify the properties of neuronal networks selectively.
[150] D. Wolpert and M. Kawato. Multiple paired forward and inverse models for motor control. Neural Networks, 11:1317-1329, 1999.
[ bib ]
[151] S. Amari. Natural gradients work efficiently in learning. Neural Computation, 10, 1998.
[ bib ]
[152] R. C. Arkin. Behavior Based Robotics. MIT Press,Cambridge, MA, 1998.
[ bib ]
[153] L. Arnold. Random Dynamical Systems. Springer, 1998.
[ bib ]
[154] E. Bienenstock and D. Lehmann. Regulated criticality in the brain? Advances in Complex Systems, 1:361-384, 1998.
[ bib ]

We propose that a regulation mechanism based on Hebbian covariance plasticity may cause the brain to operate near criticality. We analyze the effect of such a regulation on the dynamics of a network with excitatory and inhibitory neurons and uniform connectivity within and across the two populations. We show that, under broad conditions, the system converges to a critical state lying at the common boundary of three regions in parameter space; these correspond to three modes of behavior: high activity, low activity, oscillation.
[155] L. O. Chua. CNN: A paradigm for complexity. World Scientific, 1998.
[ bib ]
[156] Y. Frégnac. Homeostasis or synaptic plasticity. Nature, 391:845-846, 1998.
[ bib ]
[157] A. Goswami, B. Thuilot, and B. Espiau. A study of the passive gait of a compass-like biped robot: Symmetry and chaos. International Journal of Robotics Research, 1998.
[ bib ]
[158] W. Maass and A. Zador. Computing and learning with dynamic synapses. In W. Maass and C. Bishop, editors, Pulsed Neural Networks, pages 321-336. MIT-Press (Cambridge), 1998.
[ bib ]
[159] K.-R. Müller, A. Smola, G. Rätsch, B. Schölkopf, J. Kohlmorgen, and V. Vapnik. Using support vector machines for time series prediction. In B. Schölkopf, C. Burges, and A. Smola, editors, Advances in Kernel Methods - Support Vector Learning. MIT Press, Cambridge, MA, 1998. to appear.
[ bib ]
[160] P. Niyogi. The informational complexity of learning. Kluwer Academc Publisher, Boston/Dordrecht/London, 1998.
[ bib ]
[161] R. S. Sutton. Reinforcement Learning: An Introduction. MIT Press/Bradford Books, 1998.
[ bib ]
[162] M. M. Williamson. Neural control of rhythmic arm movements. Neural Networks, 11:1379-1394, 1998.
[ bib ]
[163] D. M. Wolpert, R. C. Miall, and M. Kawato. Internal models in the cerebellum. Trends in Cognitive Sciences, 2:338 - 347, 1998.
[ bib ]
[164] D. M. Wolpert, R. C. Miall, and M. Kawato. Internal models in the cerebellum. Trends in Cognitive Sciences, 2(9), 1998.
[ bib ]
[165] T. Ziemke. Adaptive behavior in autonomous agents. Presence: Teleoperators and Virtual Environments, 7:564-587, 1998.
[ bib ]

This paper provides an overview of the bottom-up approach to AI, commonly referred to as behavior-oriented AI. The behavior-oriented approach, with its focus on the interaction between autonomous agents and their environments, is introduced by contrasting it with the traditional approach of knowledge-based AI. Different notions of autonomy are discussed, and key problems of generating adaptive and complex behavior are identified. A number of techniques for the generation of behavior are introduced and evaluated regarding their potential for realizing different aspects of autonomy, as well as adaptivity and complexity of behavior. It is concluded that, in order to realize truly autonomous and intelligent agents, the behavior-oriented approach will have to focus even more on lifelike qualities in both agents and environments. (Author)
[166] R. D. Beer, R. D. Quinn, H. J. Chiel, and R. E. Ritzmann. Biologically inspired approaches to robotics. Communications of the ACM, 40(3):30-38, Mar. 1997.
[ bib ]
[167] R. D. Beer, R. D. Quinn, H. J. Chiel, and R. E. Ritzmann. Biologically inspired approaches to robotics: what can we learn from insects? Commun. ACM, 40(3):30-38, 1997.
[ bib ]
[168] J. Baxter. The canonical distortion measure for vector quantization and function approximation. In Proc. 14th International Conference on Machine Learning, pages 39-47. Morgan Kaufmann, 1997.
[ bib ]
[169] R. D. Beer, R. D. Quinn, H. J. Chiel, and R. E. Ritzmann. Biologically inspired approaches to robotics: what can we learn from insects? Commun. ACM, 40(3):30-38, 1997.
[ bib ]
[170] BOREL. Boulder reinforcement learning (borel) group, University of Colorado, Boulder: Java demonstrations of reinforcement learning. http://www.cs.colorado.edu/~baveja/borel.html, 1997.
[ bib ]
[171] T. G. Dietterich and N. S. Flann. Explanation-based and reinforcement learning: A unified view. Machine Learning, 28(2/3):169-210, 1997.
[ bib | .ps.gz ]
[172] D. Floreano and S. Nolfi. God save the red queen! competition in co-evolutionary robotics. In J. R. Koza, K. Deb, M. Dorigo, D. B. Fogel, M. Garzon, H. Iba, and R. L. Riolo, editors, Genetic Programming 1997: Proceedings of the Second Annual Conference, Stanford University, CA, USA, 13-16 July 1997. Morgan Kaufmann.
[ bib ]

Keywords: Artifical life and evolutionary robotics
[173] J. Fürnkranz. Pruning algorithms for rule learning. Machine Learning, 27:139, 1997.
[ bib ]
[174] D. Golomb, J. Hertz, S. Panzeri, A. Treves, and B. Richmond. How well can we estimate the information carried in neuronal responses from limited samples? Neural Computation, 9(3):649-665, 1997.
[ bib ]
[175] D. T. Hau and E. W. Coiera. Learning qualitative models of dynamic systems. Machine Learning, 26:177-211, 1997.
[ bib ]

The automated construction of dynamic system models is an important application area for ILP. We describe a method that learns qualitative models from time-varying physiological signals. The goal is to understand the complexity of the learning task when faced with numerical data, what signal processing techniques are required, and how this affects learning. The qualitative representation is based on Kuipers' QSIM. The learning algorithm for model construction is based on Coiera's GENMODEL. We show that QSIM models are efficiently PAC learnable from positive examples only, and that GENMODEL is an ILP algorithm for efficiently constructing a QSIM model. We describe both GENMODEL which performs RLGG on qualitative states to learn a QSIM model, and the front-end processing and segmenting stages that transform a signal into a set of qualitative states. Next we describe results of experiments on data from six cardiac bypass patients. Useful models were obtained, representing both normal and abnormal physiological states. Model variation across time and across different levels of temporal abstraction and fault tolerance is explored. The assumption made by many previous workers that the abstraction of examples from data can be separated from the learning task is not supported by this study. Firstly, the effects of noise in the numerical data manifest themselves in the qualitative examples. Secondly, the models learned are directly dependent on the initial qualitative abstraction chosen.

Keywords: SIMULATION, inductive logic programming, qualitative modelling, system identification, PAC learning, physiological modelling, cardiovascular system, data mining, patient monitoring
[176] Hofmann and Buhmann. Correction to ``pairwise data clustering by deterministic annealing''. IEEETPAMI: IEEE Transactions on Pattern Analysis and Machine Intelligence, 19, 1997.
[ bib ]
[177] F. C. Hoppensteadt and E. M. Izhikevich. Weakly Coupled Neural Networks. Springer, 1997.
[ bib ]
[178] C. Jacob. Principia Evolvica - Simulierte Evolution mit Mathematica. dpunkt-Verlag, Heidelberg, 1997.
[ bib ]
[179] A. Krogh and P. Sollich. Statistical mechanics of ensemble learning. Physical Review, E 55:811, 1997.
[ bib ]
[180] T.-S. Lim, W.-Y. Loh, and Y.-S. Shih. An empirical comparison of decision trees and other classification methods. Technical Report 979, Department of Statistics, University of Wisconsin-Madison, Madison, WI, June 30 1997.
[ bib | .ps.gz ]
[181] E. Malthouse, A. Tamhane, and R. Mah. Nonlinear partial least squares. Computers in Chemical Engineering, 12(8):875-890, 1997.
[ bib ]
[182] D. McFarland and E. Spier. Basic cycles, utility and opportunism in self-sufficient mobile robots. Robotics and Autonomous Systems, (20):179-190, 1997.
[ bib ]
[183] K.-R. Müller, A. Smola, G. Rätsch, B. Schölkopf, J. Kohlmorgen, and V. Vapnik. Using support vector machines for time series prediction. In W. Gerstner, A. Germond, M. Hasler, and J.-D. Nicoud, editors, Proceedings ICANN'97, Int. Conf. on Artificial Neural Networks, volume 1327 of LNCS, pages 999-1004, Berlin, 1997. Springer.
[ bib ]
[184] S. Oore, G. E. Hinton, and G. Dudek. A mobile robot that learns its place. Neural Computation, 9(3):683-699, 1997.
[ bib ]
[185] S. Nolfi. Using emergent modularity to develop control system for mobile robots. Adaptive Behavior,Special Issue on Environment Structure and Behavior, 3-4:343-364, 1997.
[ bib ]
[186] P. Nordin. Evolutionary Program induction of binary machine code and its applications. Krehl Verlag, Münster, 1997.
[ bib ]
[187] A demonstration of simulated annealing. Internet: http://www.taygeta.com/annealing/demo1.html, 1997.
[ bib ]
[188] S. Singh, P. Norvig, and D. Cohn. How to make software agents do the right thing: an introduction to reinforcement learning. http://www.cs.colorado.edu/~baveja/RLMasses/RL.html, 1997.
[ bib ]
[189] A. Steinhage. Dynamical Systems for the Generation of Navigation Behavior. PhD thesis, Ruhr-Universit at Bochum, Germany, 1997.
[ bib ]
[190] V. N. Vapnik. The support vector method. Lecture Notes in Computer Science, 1327:263-??, 1997.
[ bib ]
[191] G. von Randow. Roboter - unsere nä chsten Verwandten. Rowohlt, 1997.
[ bib ]
[192] M. J. Mataric and D. Cliff. Challenges in evolving controllers for physical robots. Journal of Robotics and Autonomous Systems, 19(1):67-83, Oct. 1996.
[ bib ]

Feasibility of applying evolutionary methods to automatically generating controllers for physical mobile robots. Overview state of the art, main approaches, key challenges, unanswered problems, promising directions

Keywords: genetic algorithms, genetic programming, robots
[193] M. J. Mataric and D. Cliff. Challenges in evolving controllers for physical robots. Journal of Robotics and Autonomous Systems, 19(1):67-83, Oct. 1996.
[ bib ]

Feasibility of applying evolutionary methods to automatically generating controllers for physical mobile robots. Overview state of the art, main approaches, key challenges, unanswered problems, promising directions

Keywords: genetic algorithms, genetic programming, robots
[194] A. S. Galanopoulos and S. C. Ahalt. Codeword distribution for frequency sensitive competitive learning with one-dimensional input data. IEEE Transactions on Neural Networks, 7(3):752-756, May 1996.
[ bib ]
[195] J. Laszlo, M. van de Panne, and E. Fiume. Limit cycle control and its application to the animation of balancing and walking. In SIGGRAPH '96: Proceedings of the 23rd annual conference on Computer graphics and interactive techniques, pages 155-162. ACM Press, 1996.
[ bib ]
[196] M. Asada. Purposive behaviour acqusition for a real robot by vision based reinforcement learning. Machine Learning, 23:279 - 304, 1996.
[ bib ]
[197] M. Gell-Mann and S. LLoyd. Information measures, effective complexity and total information. Complexity, 2:44-52, 1996.
[ bib ]
[198] S. A. Kauffman. At Home in the Universe: The Search for Laws of Self-Organization and Complexity. Oxford University Press, 1996.
[ bib ]
[199] S. Luc. The origins of intelligence, 1996.
[ bib | .html ]
[200] A. Murbach. An integrative computational model for associative learning. Kognitionswissenschaft, 5:175-138, 1996.
[ bib ]
[201] S. Nolfi. Evolving nontrivial behaviours on real robots: A garbage collecting robot. Technical Report 96-04, Institute of Psychology, C.N.R. Rome, 1996.
[ bib ]
[202] J. R. Quinlan. Improved Use of Continuous Attributes in C4.5. Journal of Artificial Intelligence, 4:77-90, 1996.
[ bib ]
[203] K. Sayood. Introduction to data compression. Morgan Kaufmann, San Francisco, 1996.
[ bib ]
[204] A. Thompson, I. Harvey, and P. Husbands. Unconstrained evolution and hard consequences. Lecture Notes in Computer Science, 1062:136-??, 1996.
[ bib ]
[205] M. Mehta, J. Rissanen, and R. Agrawal. MDL-based decision tree pruning. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining (KDD'95), pages 216-221, Aug. 1995.
[ bib | .ps ]

Keywords: Data Mining, Classification, Decision-Trees, MDL
[206] S. Ben-David, N. Eiron, and E. Kushilevitz. On self-directed learning. In COLT '95: Proceedings of the eighth annual conference on Computational learning theory, pages 136-143. ACM Press, 1995.
[ bib ]
[207] R. D. Beer. A dynamical systems perspective on agent-environment interaction. Artif. Intell., 72(1-2):173-215, 1995.
[ bib ]
[208] F. Esposito, D. Malerba, and G. Semeraro. Simplifying decision trees by pruning and grafting: New results. Lecture Notes in Computer Science, 912:287-??, 1995.
[ bib ]
[209] H. R. Everett. Sensors for mobile robots. A K Peters, 1995.
[ bib ]
[210] E. V. Glasersfeld. Radical Constructivism: A Way of Knowing and Learning. Falmer, Washington DC, 1995.
[ bib ]
[211] E. Malthouse. Nonlinear partial least squares. PhD thesis, Northwestern University, Illinois, 1995.
[ bib | .html ]
[212] M. J. Mataric. Evaluation of learning performance of situated embodied agents. Lecture Notes in Computer Science, 929:579-??, 1995.
[ bib ]
[213] T. Hofmann and J. Buhmann. Multidimensional scaling and data clustering. In G. Tesauro, D. Touretzky, and T. Leen, editors, Advances in Neural Information Processing Systems, volume 7, pages 459-466. The MIT Press, 1995.
[ bib ]
[214] S. Nolfi and D. Parisi. Evolving non-trivial behaviors on real robots: An autonomous robot that picks up objects. Lecture Notes in Computer Science, 992:243, 1995.
[ bib ]
[215] B. Fritzke. Incremental learning of local linear mappings, Sept. 25 1995.
[ bib | .ps.gz ]

A new incremental network model for supervised learning is proposed. The model builds up a structure of units each of which has an associated local linear mapping (LLM). Error information obtained during training is used to determine where to insert new units whose LLMs are interpolated from their neighbors. Simulation results for several classification tasks indicate fast convergence as well as good generalization. The ability of the model to also perform function approximation is demonstrated by an example. 1 Introduction Local (or piece-wise) linear mappings (LLMs) are an economic means of describing a }wellbehaved { function f : R n ! R m . The principle is to approximate the function (which may be given by a number of input/output samples (; i) 2 R n Theta R m ) with a set of linear mappings each of which is constrained to a local region of the input space R n . LLM-based methods have been used earlier to learn the inverse kinematics of robot arms [7], for classificat...
[216] R. F. Port and T. V. Gelder, editors. Mind as Motion. The MIT Press, Cambridge, MA, 1995.
[ bib ]
[217] S. Russell and P. Norvig. Artifical Intelligence: A Modern Approach,. Prentice-Hall, Englewood Cliffs, NJ, ISBN 0-13-103805-2, 912 pp., 1995, 1995.
[ bib ]

Keywords: AI AGENT OVERVIEW VISION KNOWLEDGE-REP REACTIVE
[218] M. D. G. Schöner and C. Engels. Dynamics of behavior: Theory and applications for autonomous robot architectures. Robotics and Autonomous Systems,, 16:213-245., 1995.
[ bib ]
[219] H.-P. Schwefel. Evolution and Optimum Seeking. Wiley, New York, 1995.
[ bib ]
[220] L. Steels. Perceptually grounded meaning creation. In V. Lesser, editor, Proceedings of the First International Conference on Multi-Agent Systems. MIT Press,Cambridge, MA, 1995.
[ bib ]

The paper proposes a mechanism for the spontaneous formation of perceptually grounded meanings under the selectionist pressure of a discrimination task. The mechanism is defined formally and the results of some simulation experiments are reported.
[221] V. N. Vapnik. The nature of statistical learning theory. Springer, 1995.
[ bib ]
[222] Y. Yuan and M. J. Shaw. Induction of fuzzy decision trees. Fuzzy Sets and Systems, 69(2):125-139, 1995.
[ bib ]
[223] P. G. Zimbardo. Psychologie. Springer, 1995.
[ bib ]
[224] E. V. Siegel and K. R. McKeown. Emergent linguistic rules from inducing decision trees: disambiguating discourse clue words. In Proceedings of the Twelfth National Conference on Artificial Intelligence, Menlo Park, CA, USA, July 1994. AAAI Press.
[ bib ]

Keywords: genetic algorithms, genetic programming
[225] L. Steels. Emergent functionality in robotic agents through on-line evolution. In R. A. Brooks and P. Maes, editors, Proceedings of the 4th International Workshop on the Synthesis and Simulation of Living Systems Artificial Life IV, pages 8-16, Cambridge, MA, USA, July 1994. MIT Press,Cambridge, MA.
[ bib ]
[226] K. Sims. Evolving virtual creatures. In SIGGRAPH '94: Proceedings of the 21st annual conference on Computer graphics and interactive techniques, pages 15-22. ACM Press, 1994.
[ bib ]
[227] K. Berns. Steuerungsansätze auf der Basis neuronaler Netze für sechseinige Laufmaschinen. Dissertation DISKI 61, 1994.
[ bib ]
[228] B. Fritzke. Growing cell structures - A self-organizing network for unsupervised and supervised learning. Neural Networks, 7(9):1441-1460, 1994.
[ bib ]

Additive unsupervised learning structure. Can be extended with RBFs to get supervised learning, RBFs are positioned automatically. Good results.
[229] M. I. Jordan and R. A. Jacobs. Hierarchical mixtures of experts and the EM algorithm. Neural Computation, 6(2):181-214, 1994.
[ bib ]

We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the mixture components are generalized linear models (GLIM's).Learning is treated as a maximum likelihood problem; in particular, we present an Expectation-Maximization (EM) algorithm for adjusting the parameters of the architecture.We also develop an on-line learning algorithm in which the parameters are updated incrementally.Comparative simulation results are presented in the robot dynamics domain.
[230] D. Michie, D. J. Spiegelhalter, and C. C. Taylor. Machine learning, neural and statistical classification. Ellis Hoorwood, 1994. STATLOG-Report.
[ bib ]
[231] S. Miyakoshi, M. Yamakita, and K. Furuta. Juggling control using neural oscillators. In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 1186-1193, 1994.
[ bib ]
[232] V. Pratt. Chu spaces: Automata with quantum aspects. In Proc. Workshop on Physics and Computation (PhysComp'94), pages 186-195, Dallas, 1994. IEEE.
[ bib ]
[233] J. R. Quinlan and R. M. Cameron-Jones. Efficient top-down induction of logic programs. SIGART Bulletin, 5(1):33-42, 1994.
[ bib ]
[234] I. Rechenberg. Evolutionsstrategie'94. Frommann-Holzboog, Stuttgart, 1994.
[ bib ]
[235] D. Schlierkamp-Voosen and H. Mühlenbein. Strategy adaptation by competing subpopulations. Lecture Notes in Computer Science, 866:199-??, 1994.
[ bib ]
[236] K. Sims. Evolving virtual creatures. In SIGGRAPH '94: Proceedings of the 21st annual conference on Computer graphics and interactive techniques, pages 15-22. ACM Press, 1994.
[ bib ]
[237] L. Steels. Equilibrium analysis of behavior systems. In A. G. Cohn, editor, Proceedings of the Eleventh European Conference on Artificial Intelligence, pages 714-718, Chichester, Aug.8-12  1994. John Wiley and Sons.
[ bib ]
[238] G. Taga. Emergence of bipedal locomotion through entrainment among the neuro-musculo-skeletal system and the environment. In Proceedings of the NATO advanced research workshop and EGS topical workshop on Chaotic advection, tracer dynamics and turbulent dispersion, pages 190-208, New York, NY, USA, 1994. Elsevier North-Holland, Inc.
[ bib ]
[239] J. Wnek and R. Michalski. Hypothesis-driven constructive induction in aq17-hci: A method and experiments. Machine Learning(14):139-168, 1994.
[ bib ]
[240] R. Brooks and L. A. Stein. Building brains for bodies. Technical report, Artificial Intelligence Laboratory, Massachusetts Institute of Technology (MIT), Cambridge, Massachusetts, Aug. 1993.
[ bib ]

We describe a project to capitalize on newly available levels of computational resources in order to understand human cognition. We will build an integrated physical system including vision, sound input and output, and dextrous manipulation, all controlled by a continuously operating large scale parallel MIMD computer. The resulting system will learn to ``think'' by building on its bodily experiences to accomplish progressively more abstract tasks. Past experience suggests that in attempting to build such an integrated system we will have to fundamentally change the way artificial intelligence, cognitive science, linguistics, and philosophy think about the organization of intelligence. We expect to be able to better reconcile the theories that will be developed with current work in neuroscience.
[241] J. R. Quinlan. Combining instance-based and model-based learning. In Proceedings of the Tenth International Conference on Machine Learning, pages 236-243, Amherst, Massachusetts, July 1993. Morgan Kaufmann.
[ bib ]
[242] P. Dagum and M. Luby. Approximating probabilistic inference in bayesian belief networks is NP-hard. Artificial Intelligence, 60(1):141-153, Mar. 1993.
[ bib ]
[243] T. Bäck and H.-P. Schwefel. An Overview of Evolutionary Algorithms for Parameter Optimization. Evolutionary Computation, 1(1):1-23, 1993.
[ bib ]
[244] R. D. Beer, R. E. Ritzmann, and T. McKenna, editors. Biological Neural Networks in Invertebrate Neuroethology and Robotics. Academic Press, New York, 1993.
[ bib ]
[245] K. Berns, B. Muller, and R. Dillmann. Dynamic control of a robot leg with self-organizing feature maps. In IROS '93. Proceedings of the 1993 IEEE/RSJ International Conference on Intelligent Robots and Systems. Intelligent Robots for Flexibility (Cat. No.93CH3213-6), volume 1, pages 553-60, New York, NY, USA, 1993. IEEE.
[ bib ]
[246] I. Harvey, P. Husbands, and D. T. Cliff. Genetic convergence in a species of evolved robot control architectures. In S. Forrest, editor, Proceedings of the Fifth International Conference on Genetic Algorithms, page 636, San Mateo, CA, USA, 1993. Morgan Kaufmann. Poster version of [?].
[ bib ]
[247] I. Harvey, P. Husbands, and D. T. Cliff. Genetic convergence in a species of evolved robot control architectures. In S. Forrest, editor, Proceedings of the Fifth International Conference on Genetic Algorithms, page 636, San Mateo, CA, USA, 1993. Morgan Kaufmann. Poster version of [?].
[ bib ]
[248] P. Husbands, I. Harvey, and D. Cliff. Analysing recurrent dynamical networks evolved for robot control. Technical Report Cognitive Science Research Paper CSRP265, School of Cognitive and Computing Sciences, University of Sussex, Brighton BN1 9QH, England, UK, 1993.
[ bib ]
[249] R. A. Jacobs and M. I. Jordan. Learning piecewise control strategies in a modular neural-network architecture. IEEE Transactions on Systems, Man, and Cybernetics, 23(2):337-345, 1993.
[ bib ]

The dynamics of nonlinear systems often vary qualitatively over their parameter space. Methodologies for designing piecewise control laws for dynamical systems, such as gain scheduling, are useful because they circumvent the problem of determining a single global model of the plant dynamics. Instead, the dynamics are approximated using local models that vary with the plant's operating point. When a controller is learned instead of designed, analogous issues arise. This article describes a multi-network, or modular, neural network architecture that learns to perform control tasks using a piecewise control strategy. The architecture's networks compete to learn the training patterns. As a result, a plant's parameter space is adaptively partitioned into a number of regions, and a different network learns a control law in each region. This learning process is described in a probabilistic framework and learning algorithms that perform gradient ascent in a log likelihood function are discussed. Simulations show that the modular architecture's performance is superior to that of a single network on a multipayload robot motion control task.
[250] S. A. Kauffman. The Origins of Order: Self-Organization and Selection in Evolution. Oxford University Press, 1993.
[ bib ]
[251] J. R. Koza. Genetic Programming: On the Programming of Computers by Means of Natural Selection. 2nd edition. MIT Press, 1993.
[ bib ]
[252] J. M. Vleugels, J. N. Kok, and M. H. Overmars. Motion planning using a colored Kohonen map. Technical report, Utrecht University Technical Report RUU-CS-93-83, 1993.
[ bib ]
[253] R. C. Arkin. Homeostatic control for a mobile robot: Dynamic replanning in hazardous environments. Journal of Robotic Systems, 9(2):197-214, Mar. 1992.
[ bib ]
[254] C. Langton. Life at the edge of chaos. In C. G. Langton, C. Taylor, J. D. Farmer, and S. Rasmussen, editors, Proceedings of the Workshop on Artificial Life (ALIFE '90), volume 5 of Santa Fe Institute Studies in the Sciences of Complexity, pages 41-92, Redwood City, CA, USA, Feb. 1992. Addison-Wesley.
[ bib ]
[255] S. Arikawa, S. Kuhara, S. Miyano, Y. Mukouchi, A. Shinohara, and T. Shinohara. A machine discovery from amino acid sequences by decision trees over regular patterns. In Proceedings of the International Conference on Fifth Generation Computer Systems, pages 618-625, ICOT, Japan, 1992. Association for Computing Machinery.
[ bib ]
[256] C. K. Chui. An Introduction to Wavelets, volume 1 of Wavelet Analysis and its Applications. Academic Press, Inc., 1992.
[ bib ]

This is the first volume in the series WAVELET ANALYSIS AND ITS APPLICATIONS. It is an introductory treatise on wavelet analysis, with an emphasis on spline wavelets and and time-frequency analysis. Among the basic topics covered are time frequency localization, intergral wavelet transforms, dyadic wavelets, frames, spine wavelets, orthonormal wavelet bases, and wavelet packets. Is is suitable as a textbook for a beginning course on wavelet analysis and is directed toward both mathematicians and engineers who wish to learn about the subject.

Keywords: An Overview, Fourier Analysis, Wavelet Transforms and Time-frequency Analysis, Cardinal Spline Analysis, scaling functions and Wavelets, Cardinal Spline Wavelets, Orthogonal Wavelets and Wavelet Packets
[257] D. Cliff, P. Husbands, and I. Harvey. Analysis of evolved sensory-motor controllers. Technical Report Cognitive Science Research Paper CSRP264, School of Cognitive and Computing Sciences, University of Sussex, Brighton BN1 9QH, England, UK, 1992.
[ bib ]
[258] M. Cohen. The construction of arbitrary stable dynamics in nonlinear neural networks. Neural Networks, 5:83-103, 1992.
[ bib ]
[259] I. Daubechies. Ten Lectures on Wavelets, volume 61 of CBMS-NSF Regional Conference Series in Applied Mathematics. Society for Industrial and Applied Mathematics, Philadelphia, 1992.
[ bib | http ]
[260] A. Gersho and R. M. Gray. Vector quantization and signal processing. Kluwer Academic Publishers, Boston/Dordrecht/London, 1992.
[ bib ]
[261] F. Lari and A. Zakhor. Automatic classification of active sonar data using time-frequency transforms. In Time-Frequency and Time-Scale Analysis, pages 21-24, Victoria, B.C., Canada, 1992. IEEE Signal Processing Society.
[ bib ]

Automatic classification of active sonar signals using the Wigner-Ville transform (WVT), the wavelet transform (WT) and the scalogram is addressed. Features are extracted by integrating over regions in the time-frequency (TF) distribution, and are classified by a decision tree. Experimental results show classification and detection rates of up to 92 WVT and the scalogram, particularly at high noise levels. This can be partially attributed to the absence of cross terms in the WT.
[262] M. J. Liebmann, T. F. Edgar, and L. S. Ladson. Efficient data reconciliation and estimation for dynamic processes using nonlinear dynamic programming techniques. Computers in Chemical Engineering, 16(10/11):961 - 985, 1992.
[ bib ]
[263] C. J. Watkins and P. Dayan. Technical note Q-learning. Machine Learning, 8:279, 1992.
[ bib ]
[264] K. D. Miller and D. J. C. MacKay. The role of constraints in Hebbian learning. Technical Report CNS Memo 19, Pasadena, California, 1992.
[ bib ]
[265] A. E. Nicholson and J. M. Brady. The data association problem when monitoring robot vehicles using dynamic belief networks. In ECAI 92: 10th European Conference on Artificial Intelligence Proceedings, pages 689-693, Vienna, Austria, 3-7 Aug. 1992. Wiley.
[ bib ]
[266] J. R. Quinlan. C4.5: Programs for Machine Learning. Morgan Kaufmann, 1992.
[ bib ]

Book only: ISBN 1-55860-238-0 $ 44.95 U.S. $49.45 International Book and Software: ISBN 1-55860-240-2 $69.95 U.S. $76.00 International Software only: ISBN 1-55860-239-9 $34.95 U.S. $38.45 International 1 Introduction 2 Constructing Decision Trees 3 Unknown Attribute Values 4 Pruning Decision Trees 5 From Trees to Rules 6 Windowing 7 Grouping Attribute Values 8 Interacting with Classification Models 9 Guide to Using the System 10 Limitations 11 Desirable Additions

Keywords: book, decision trees, C4.5, ID3, inductive inference, II, rule rules
[267] R. J. Williams. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine Learning, 8:229 - 256, 1992.
[ bib ]
[268] C. Schaffer. When does overfitting decrease prediction accuracy in induced decision trees and rule sets? In Y. Kodratoff, editor, Proceedings of the European Working Session on Learning : Machine Learning (EWSL-91), volume 482 of LNAI, pages 192-205, Porto, Portugal, Mar. 1991. Springer Verlag.
[ bib ]
[269] W. Bechtel and A. Abrahamsen. Connectionism and the Mind: An Introduction to Parallel Processing in Networks,. Basil Blackwell, Cambridge, MA:, 1991.
[ bib ]
[270] R. D. Beer and H. J. Chiel. The neural basis of behavioral choice in an artifical insect. In J.-A. Meyer and S. W. Wilson, editors, From animals to animats, pages 247-254. First International Conference on Simulation of Adaptive Behavior, 1991.
[ bib ]
[271] E. Bloedorn and R. S. Michalski. Data-driven constructive induction in AQ17-PRE: A method and experiments. In Proc. Third International Conference on Tools for Artificial Intelligence, pages 30-37, Los Alamitos, CA, 1991. IEEE Computer Society Press.
[ bib ]
[272] R. Dillmann. Informationsverarbeitung in der Robotik. Springer, 1991.
[ bib ]
[273] F. Hoffmeister and T. Bäck. Genetic algorithms and evolution strategies: similarities and differences. In H. P. Schwefel and R. Männer, editors, Parallel Problem Solving from Nature - Proceedings of 1st Workshop, PPSN 1, volume 496 of Lecture Notes in Computer Science, pages 455-469, Dortmund, Germany, 1-3 Oct. 1991. Springer-Verlag, Berlin, Germany.
[ bib ]
[274] R. A. Jacobs, M. I. Jordan, S. J. Nowlan, and G. E. Hinton. Adaptive mixtures of local experts. Neural Computation, 3(1):79-87, 1991.
[ bib ]
[275] J. R. Koza. Concept formation and decision tree induction using the genetic programming paradigm. Lecture Notes in Computer Science, 496:124-??, 1991.
[ bib ]
[276] H. Mühlenbein. Darwin's continent cycle theory and its simulation by the prisoner's dilemma. Complex Systems, 5:459-478, 1991.
[ bib ]
[277] P. Quinlan. Connectionism and Psychology. Harvester Wheatsheaf, New York, 1991.
[ bib ]
[278] R. S. Sutton. Reinforcement learning architectures for animats. In Proceedings of th International Workshop on the Simulation of Adaptive Behavior: From Animals to Animats, pages 288-296, Cambridge, Massachusetts, 1991. MIT Press.
[ bib ]
[279] R. S. Sutton. Integrated architectures for learning, planning and reacting based on approximating dynamic programming. In Proceedings of the Seventh International Conference on Machine Learning, June 1990.
[ bib ]
[280] R. Brooks. The behavior language: User's guide. Technical report, Artificial Intelligence Laboratory, Massachusetts Institute of Technology (MIT), Cambridge, Massachusetts, Apr. 1990.
[ bib ]

The Behavior Language is a rule-based real-time parallel robot programming language originally based on ideas from [Brooks 86], [Connell 89], and [Maes 89]. It compiles into a modified and extended version of the subsumption architecture [Brooks 86] and thus has backends for a number of processors including the Motorola 68000 and 68HC11, the Hitachi 6301, and Common Lisp. Behaviors are groups of rules which are activatable by a number of different schemes. There are no shared data structures across behaviors, but instead all communication is by explicit message passing. All rules are assumed to run in parallel and asynchronously. It includes the earlier notions of inhibition and suppression, along with a number of mechanisms for spreading of activation.
[281] S. C. Ahalt, A. K. Krishnamurthy, P. Chen, and D. E. Melton. Competitive learning algorithms for vector quantization. Neural Networks, 3:277-290, 1990.
[ bib ]

Keywords: vector quantization | speech | competitive learning
[282] A. G. Barto, R. S. Sutton, and C. Watkins. Learning and sequential decision making. In M. Gabriel and J. Moore, editors, Learning and computational neuroscience : foundations of adaptive networks. M.I.T. Press, Cambridge, Mass, 1990. Interent: ftp://ftp.cs.umass.edu/pub/anw/pub/sutton/barto-sutton-watkins-90.ps.gz.
[ bib ]
[283] R. D. Beer, H. J. Chiel, and L. S. Sterling. A biological perspective on autonomous agent design. In P. Maes, editor, Designing Autonomous Agents, pages 169-185. MIT Press,Cambridge, MA, 1990.
[ bib ]
[284] C. W. Gardiner. Handbook of Stochastic Methods. Springer, 1990.
[ bib ]
[285] J. Honerkamp. Stochastische dynamische Systeme. VCH Verlagsgesellschaft, 1990.
[ bib ]
[286] C. G. Langton. Computation at the edge of chaos. Physica D, 42, 1990.
[ bib ]
[287] T. McGeer. Passive dynamic walking. International Journal of Robotics Research, 9, No., 2,:62-82, 1990.
[ bib ]
[288] D. Nguyen and B. Widrow. The truck backer-upper: An example of self-learning in neural networks. In W. T. Miller, R. S. Sutton, and P. J. Werbos, editors, Neural networks for control, chapter 12. M.I.T. Press, Cambridge, Mass, 1990.
[ bib ]
[289] K. Rose, E. Gurewitz, and G. C. Fox. Statistical mechanics and phase transitions in clustering. Technical Report CFP-895, California Institute of Technology, 1990.
[ bib ]
[290] H. P. Schwefel. Parallel problem solving from nature. Springer, Berlin [u.a.], 1990.
[ bib ]
[291] J. W. Shavlik and T. G. Dietterich, editors. Readings in Machine Learning. Morgan Kaufmann, 1990.
[ bib ]
[292] R. S. Sutton and A. G. Barto. Time-derivative models of Pavlovian reinforcement. In J. W. Moore and M. Gabriel, editors, Learning and Computational Neuroscience. MIT Press, 1990.
[ bib ]
[293] P. J. Werbos. A menu of designs for reinforcement learning over time. In W. T. Miller, R. S. Sutton, and P. J. Werbos, editors, Neural networks for control, chapter 3, pages 67-95. M.I.T. Press, Cambridge, Mass, 1990.
[ bib ]
[294] J. C. Craig. Introduction to Robotics, 2nd Edition. Addison Wesley, 1989. 2nd Edition 1991.
[ bib ]
[295] D. E. Goldberg. Genetic Algorithms in Search, Optimization, and Machine Learning. Addison Wesley, 1989.
[ bib ]
[296] B. Libet. The timing of a subjective experience. Behavioral and Brain Sciences, 12:183-185, 1989.
[ bib ]
[297] P. Prusinkiewicz and J. S. Hanan. Lindenmayer Systems, Fractals, and Plants. Springer, New York, NY, 1989.
[ bib ]
[298] H. Risken. The Fokker-Planck equation. Springer, 1989.
[ bib ]
[299] G. Salton. Automatic Text Processing. Addison Wesley, Massachusetts, 1989.
[ bib ]
[300] K. Binder and D. W. Heerman. Monte Carlo Simulation in Statistical Mechanics. Springer-Verlag, Berlin, 1988.
[ bib ]
[301] G. Kampis. Information, computation and complexity. In M. Carvallo, editor, Nature, Cognition and Systems, pages 313-320. Kluer, Dordrecht, 1988.
[ bib ]
[302] R. S. Sutton. Learning to predict by the methods of temporal differences. Machine Learning, 3:9-44, 1988.
[ bib ]
[303] R. Brooks. Intelligence without representation. In Workshop on Foundations of AI, June 1987.
[ bib ]
[304] K. S. Fu and R. C. Gonzalez. Robotics. McGraw-Hill, 1987.
[ bib ]
[305] H. Haken. Advanced Synergetics. Springer, Berlin, 1987.
[ bib ]
[306] G. Kampis and V. Csanyi. Notes on order and complexity. Journal of Theoretical Biology, 124:111-121, 1987.
[ bib ]
[307] R. Brooks. Achieving artificial intelligence through building robots. Technical report, Artificial Intelligence Laboratory, Massachusetts Institute of Technology (MIT), Cambridge, Massachusetts, May 1986.
[ bib ]

We argue that generally accepted methodologies of artificial intelligence research are limited in the proportion of human level intelligence they can be expected to emulate. We argue that the currently accepted decompositions and static representations used in such research are wrong. We argue for a shift to a process based model, with a decomposition based on task achieving behaviors as the organizational principle. In particular we advocate building robotic insects.
[308] P. R. Cohen and E. W. Feigenbaum. The Handbook of Artificial Intelligence Vol. 3. Addison Wesley, 1986.
[ bib ]
[309] J. R. Quinlan. Induction of decision trees. Machine Learning, 1:81-106, 1986.
[ bib ]
[310] C. W. Gardiner. Handbook of Stochastic Methods. Springer, 1985.
[ bib ]
[311] G. Kampis and V. Csanyi. Simple models do not eliminate complexity from the real world. Journal of Theoretical Biology, 115:467-469, 1985.
[ bib ]
[312] B. Libet. Unconscious cerebral initiative and the role of conscious will in voluntary action. Behavioral and Brain Sciences, 8:529-566, 1985.
[ bib ]
[313] K. Matsuoka. Sustained oscillations generated by mutually inhibiting neurons with adaptation. Biol. Cybern., 52:367-376, 1985.
[ bib ]
[314] V. Braitenberg. Vehicles: Experiments in Synthetic Psychology. MIT Press, 1984.
[ bib ]
[315] J. R. QUINLAN. Inductive inference as a tool for the construction of efficient classification programs. Tioga, Palo Alto, CA, 1983.
[ bib ]
[316] W. Ebeling and R. Faistel. Physik der Selbstorganisation und Evolution. Akademie-Verlag, Berlin, 1982.
[ bib ]
[317] N. G. van Kampen. Stochastic Processes in Physics and Chemistry. Elsevier, 1981.
[ bib ]
[318] A. Feldman. Superposition of motor programs, I. Rhytmic forearm movements in man. Neuroscience, 5:81-90, 1980.
[ bib ]
[319] Y. Linde, A. Buzo, and R. M. Gray. An algorithm for vector quantizer design. IEEE Transactions Comm., 28:84-95, 1980.
[ bib ]
[320] R. A. Rescorla. Pavlovian Second-Order Conditioning: Studies in Associative Learning. Erlbaum/Wiley, 1980.
[ bib ]
[321] H. R. Maturana and F. J. Varela. Autopoiesis and Cognition: The Realization of the Living. Reidel, Boston, 1979.
[ bib ]
[322] W. N. Wapnik and A. J. Tscherwonenkis. Theorie der Zeichenerkennung. Akademie-Verlag, 1979.
[ bib ]
[323] I. Rechenberg. Evolutionsstrategie: Optimierung technischer Systeme nach Prinzipien der biologischen Evolution. Frommann-Holzboog, Stuttgart, 1973.
[ bib ]
[324] R. O. Duda and P. E. Hart. Pattern Classification and Scene Analysis. Wiley, New York, 1972.
[ bib ]
[325] R. A. Rescorla and A. R. Wagner. A theory of pavlovian conditioning: variations in the effectiveness of reinforcement and nonreinforcement. In A. H. Black and W. F. Prokasy, editors, Classical Conditioning II: Current Research and Theory, pages 64 - 99. Appleton-Century-Crofts, 1972.
[ bib ]
[326] A. Lindenmayer. Developmental systems without cellular interactions, their languages and grammars. Journal of Theoretical Biology, 30:455-484, 1971.
[ bib ]
[327] A. Lindenmayer. Mathematical models for cellular interactions in development, I. filaments with one-sided inputs. Journal of Theoretical Biology, 18:280-299, 1968.
[ bib ]
[328] A. Lindenmayer. Mathematical models for cellular interactions in development, II. simple and branching filaments with two-sided inputs. Journal of Theoretical Biology, 18:300-315, 1968.
[ bib ]
[329] N. A. Bernstein. The Co-Ordination and Regulation of Movements. Pergamon Press, 1967.
[ bib ]
[330] W. R.Ashby. Design for a Brain. Chapman and Hill, London, 1954.
[ bib ]
[331] N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller. J. Chem. Phys. 21, 21:1087, 1953.
[ bib ]
[332] W. B. Cannon. The Wisdom of the Body. Norton, New York, 1939.
[ bib ]
[333] E. C. Tolman. Purposive Behavior in Animals and Men. Appleton-Century-Crofts, 1932.
[ bib ]
[334] H. Poincaré. La Valeur de la Science. Flammarion, Paris, 1905.
[ bib ]
[335] G. Berkeley. An Essay Towards a New Theory of Vision. 1709.
[ bib ]
[336] M. Gutknecht, R. Pfeifer, and M. Stolze. Cooperative hybrid systems. In Artificial intelligence, IJCAI-91, Proc. 12th Int. Conf., Sydney/Australia 1991, 824-829 .
[ bib ]

[For the entire collection see Zbl. 741.68016.]

The authors' effort to build a situated knowledge-based system resultet in a cooperative hybrid system called DDT ( device diagnostic tool). DDT is a hybrid symbolic/connectionist system, embodying cooperativity and self-tuning capabilities, thus being able to face the problem of model explosion cycle. The approach is illustrated using a real-life expert system in the domain of technical troubleshooting.

Keywords: situated knowledge-based system; device diagnostic tool
[337] L. e. Steels, G. e. Schreiber, and W. e. van de Velde. A future for knowledge acquisition. 8th European knowledge acquisition workshop, EKAW '94, Hoegaarden, Belgium, September 26-29, 1994. Proceedings. Lecture Notes in Computer Science. Lecture Notes in Artificial Intelligence. 867. Berlin: Springer-Verlag. xii, 413 p. DM 80.00; öS 624.00; sFr 80.00 .
[ bib ]

The articles of this volume will be reviewed individually.

Keywords: Hoegaarden (Belgium); Workshop; Proceedings; EKAW '94; Knowledge acquisition
[338] L. Steels. The origins of syntax in visually grounded robotic agents. Artif. Intell., 103(1-2):133-156.
[ bib ]

The paper proposes a set of principles and a general architecture that may explain how language and meaning may originate and complexify in a group of physically grounded distributed agents. An experimental setup is introduced for concretising and validating specific mechanisms based on these principles. The setup consists of two robotic heads that watch static or dynamic scenes and engage in language games, in which one robot describes to the other what they see. The first results from experiments showing the emergence of distinctions, of a lexicon, and of primitive syntactic structures are reported.

Keywords: symbolic models; robotic agents; origins of language and meaning
[339] D.-H. Kim and J.-H. Kim. A real-time limit-cycle navigation method for fast mobile robots and its application to robot soccer. Robot. Auton. Syst., 42(1):17-30.
[ bib ]

Summary: A mobile robot should be designed to navigate with collision avoidance capability in the real world, flexibly coping with the changing environment. In this paper, a novel limit-cycle navigation method is proposed for a fast mobile robot using the limit-cycle characteristics of a 2nd-order nonlinear function. It can be applied to the robot operating in a dynamically changing environment, such as in a robot soccer system. By adjusting the radius of the motion circle and the direction of obstacle avoidance, the navigation method proposed enables a robot to maneuver smoothly towards any desired destination. Simulations and real experiments using a robot soccer system demonstrate the merits and practical applicability of the proposed method.

Keywords: robot soccer; navigation; limit cycle; real-time control
[340] L. Lichtensteiger and R. Pfeifer. An optimal sensor morphology improves adaptability of neural network controllers. In Dorronsoro, José R. (ed.), Artificial neural networks - ICANN 2002. 12th international conference, Madrid, Spain, August 28-30, 2002. Proceedings. Berlin: Springer. Lect. Notes Comput. Sci. 2415, 850-855 .
[ bib ]
[341] L. Steels. The evolution of communication systems by adaptive agents. In Alonso, Eduardo (ed.) et al., Adaptive agents and multi-agent systems. Adaptation and multi-agent learning. Berlin: Springer. Lect. Notes Comput. Sci. 2636, 125-140 .
[ bib ]

Summary: The paper surveys some of the mechanisms that have been demonstrated to be relevant for evolving communication systems in software simulations or robotic experiments. In each case, precursors or parallels with work in the study of artificial life and adaptive behaviour are discussed.
[342] R. te Boekhorst, M. Lungarella, and R. Pfeifer. Dimensionality reduction through sensory-motor coordination. In Kaynak, Okyay (ed.) et al., Artificial neural networks and neural information processing - ICANN/ICONIP 2003. Joint international conference ICANN/ICONIP 2003, Istanbul, Turkey, 26-29, 2003. Proceedings. Berlin: Springer. Lect. Notes Comput. Sci. 2714, 496-503 .
[ bib ]

Summary: The problem of category learning has been traditionally investigated by employing disembodied categorization models. One of the basic tenets of embodied cognitive science states that categorization can be interpreted as a process of sensory-motor coordination, in which an embodied agent, while interacting with its environment, can structure its own input space for the purpose of learning about categories. Many researchers, including John Dewey and Jean Piaget, have argued that sensory-motor coordination is crucial for perception and for development. In this paper we give a quantitative account of why sensory-motor coordination is important for perception and category learning.
[343] L. Steels and J.-C. Baillie. Shared grounding of event descriptions by autonomous robots. Robotics and Autonomous Systems., 43(2-3):163-173.
[ bib ]

Summary: The paper describes a system for open-ended communication by autonomous robots about event descriptions anchored in reality through the robot's sensori-motor apparatus. The events are dynamic and agents must continually track changing situations at multiple levels of detail through their vision system. We are specifically concerned with the question how grounding can become shared through the use of external (symbolic) representations, such as natural language expressions.

Keywords: Autonomous Robots; Event Descriptions; Open-Ended
[344] J. Tani. Learning to generate articulated behavior through the bottom-up and the top-down interaction processes. Neural Networks. The Official Journal of the International Neural Network Society, European Neural Network Society, Japanese Neural Network Society., 16(1):11-23.
[ bib ]

Summary: A novel hierarchical neural network architecture for sensory-motor learning and behavior generation is proposed. Two levels of forward model neural networks are operated on different time scales while parametric interactions are allowed between the two network levels in the bottom-up and top-down directions. The models are examined through experiments of behavior learning and generation using a real robot arm equipped with a vision system. The results of the learning experiments showed that the behavioral patterns are learned by self-organizing the behavioral primitives in the lower level and combining the primitives sequentially in the higher level. The results contrast with prior work by Pawelzik et al. [Neural Comput. 8, 340 (1996)], Tani and Nolfi [From animals to animats, 1998], and Wolpert and Kawato [Neural Networks 11, 1317 (1998)] in that the primitives are represented in a distributed manner in the network in the present scheme whereas, in the prior work, the primitives were localized in specific modules in the network. Further experiments of on-line planning showed that the behavior could be generated robustly against a background of real world noise while the behavior plans could be modified flexibly in response to changes in the environment. It is concluded that the interaction between the bottom-up process of recalling the past and the top-down process of predicting the future enables both robust and flexible situated behavior.

Keywords: Learning; Behavior; Articulation; Chunking
[345] R. Pfeifer. Embodied Artificial Intelligence - On the role of morphology and materials in the emergence of cognition. In Schubert, Sigrid E.; Reusch, Bernd; Jesse, Norbert (eds.). Informatik bewegt: Informatik 2002 - 32. Jahrestagung der Gesellschaft für Informatik e.v. (GI), 30. September - 3.Oktober 2002 in Dortmund. Bonn: Köllen Verlag. Lect. Notes Inform. 19, p. 49-63 .
[ bib ]

Entire collection CS 2004.019000G
[346] L. Steels. Equilibrium analysis of behavior systems. In Cohn, A. G. (ed.). 11th European Conference on Artificial Intelligence. Proceedings. (ECAI '94, Amsterdam, NL, Aug.8-12, 1994). Chichester: John Wiley & Sons. 832 p., 714-718 p. .
[ bib ]

A behavior system consists of the components and control programs necessary to establish a particular behavior in a robotic agent. The paper proposes a mathematical approach for the analysis of behavior systems. The approach rests on viewing a behavior system as a dynamical system whose equilibrium state is attained when the behavior it is responsible for is achieved.

Keywords: behavior system; robotic agent; dynamical system; equilibrium state; autonomous agent
[347] J. Tani. Self-organizing of symbolic processes through interaction with the physical world. In Mellish, Chris S. (ed.). IJCAI-95. 14th international joint conference on artificial intelligence. Proceedings. (IJCAI-95. 14th International Joint Conference on Artificial Intelligence, Montreal (Canada), 20-25 Aug 1995). San Mateao, CA: Morgan Kaufmann Publishers, Inc.. XLIV, 2077 p., p. 112-118 .
[ bib ]

The paper describes how symbolic processes are self-organized in the navigational learning of a mobile robot. Based on a dynamical system's approach, the paper shows that the forward modeling scheme based on recurrent neural network (RNN) learning is capable of extracting grammatical structure hidden in the geometry of the workspace from navigational experience. This robot is capable of mentally simulating its own actions using the acquired forward model. The paper shows that such a mental process by the RNN can naturally be situated with respect to the behavioural contexts, provided that the forward model learned is embedded on the global attractor. The internal representation obtained is proved to be grounded, since it is self-organized solely through itnteraction with the physical world. The paper shows also that structural stability arises in the interaction between the neural dynamics and the environment dynamics, accounting for the situatedness of the internal symbolic process.

Keywords: self-organization of symbolic processes; robot navigation; situatedness of symbolic processes
[348] J. Tani and N. Fukumura. Self-organizing internal representation in learning of navigation: a physical experiment by the mobile robot YAMBICO.
[ bib ]

The paper discusses a novel scheme for sensory-based navigation of a mobile robot. In our previous work (Tani & Fukumura, 1994, Neural Networks, 7(3), 553-563), we formulated the problem of goal-directed navigation as an embedding problem of dynamical systems: desired trajectories in a task space should be embedded in an adequate sensory-based internal state space so that a unique mapping from the internal state space to the motor command could be established. In the current formulation a recurrent neural network is employed, which shows that an adequate internal state space can be self-organized, through supervised training with sensorimotor sequences. The experiment was conducted using a real mobile robot equipped with a laser range sensor, demonstrating the validity of the presented scheme by working in a noisy real-world environment.

Keywords: neural networks; sensory motor system; mobile robot; navigation learning
[349] N. Franceschini and R. Chagneux. Repetitive scanning in the fly compound eye. In Proceedings Gottingen Neurobiology Conference}.
[ bib ]
[350] R. Isaacs. Differential Games.
[ bib ]
[351] T. K. Leen. Stochastic Manhattan learning: An exact time-evolution operator for the ensemble dynamics. ftp://speech.cse.ogi.edu/pub/neural/papers/LeenMoody96.StochMan.ps.Z.
[ bib ]
[352] L. Moreau, E. Sontag, and M. Arcak. Automated tuning of bifurcations via feedback. to be published.
[ bib ]
[353] Learning decision trees for mapping the local environment in mobile robot navigation. Technical report.
[ bib | .ps.gz ]

This paper describes the use of the C4.5 decision tree learning algorithm in the design of a classifier for a new approach to the mapping of a mobile robot's local environment. The decision tree uses the features from the echoes of an ultrasonic array mounted on the robot to classify the contours of its local environment. The contours are classified into a finite number of two dimensional shapes to form a primitive map which is to be used for navigation. The nature of the problem, noise and the practical timing constraints, distinguishes it from those typically used in machine learning applications and highlights some of the advantages of decision tree learning in robotic applications.
[354] L. Righetti, J. Buchli, and A. Ijspeert. Dynamic hebbian learning in adaptive frequency oscillators. Physica D. In Press.
[ bib | .pdf ]
[355] Test. Test, test. test.
[ bib ]
[356] R. Der. Vorlesung Maschinelles Lernen. UniversitätLeipzig Institut für Informatik. kap. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/.
[ bib ]
[357] R. Der. Vorlesung Maschinelles Lernen. UniversitätLeipzig Institut für Informatik. kap.1: Einführung. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/ml-intr.ps.Z.
[ bib ]
[358] R. Der. Vorlesung Maschinelles Lernen. UniversitätLeipzig Institut für Informatik. kap. 2: Parameteradaptive Lernverfahren. http://www.informatik.uni-leipzig.de/{~}der/Vorlesungen/ml-parad.ps.Z.
[ bib ]
[359] R. Der. Vorlesung Maschinelles Lernen. UniversitätLeipzig Institut für Informatik. kap. 3: >klassifikationslernen. http://www.informatik.uni-leipzig.de/{~}der/Vorlesungen/ml-klass.ps.Z.
[ bib ]
[360] R. Der. Vorlesung Neuroinformatik. UniversitätLeipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/skrpt-ni.html.
[ bib ]
[361] R. Der. Vorlesung Neuroinformatik. Kapitel 1. UniversitätLeipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/grundlgn.ps.Z.
[ bib ]
[362] R. Der. Vorlesung Neuroinformatik. Kapitel 2. UniversitätLeipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/el-proz.ps.Z.
[ bib ]
[363] R. Der. Vorlesung Neuroinformatik. Kapitel 3. UniversitätLeipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/ni-feed.ps.Z.
[ bib ]
[364] R. Der. Vorlesung Neuroinformatik. Kapitel 4. UniversitätLeipzig, Institut für Informatik.
[ bib ]
[365] R. Der. Vorlesung Neuroinformatik. Kapitel 5. UniversitätLeipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/ni-zeit.ps.Z.
[ bib ]
[366] R. Der. Vorlesung Neuroinformatik. Kapitel 6. UniversitätLeipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/ni-attraktor.ps.Z.
[ bib ]
[367] R. Der. Vorlesung Neuroinformatik. Kapitel 7. . Universität Leipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/ni-som.ps.Z.
[ bib ]
[368] R. Der. Vorlesung Neuroinformatik. Kapitel 8. Universität Leipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/ni-automat.ps.Z.
[ bib ]
[369] R. Der. Vorlesung Neuroinformatik. Kapitel 9: Neuro-fuzzy verfahren. Universität Leipzig, Institut für Informatik.
[ bib ]
[370] R. Der. Vorlesung Robotik. UniversitätLeipzig Institut für Informatik. kap. 2.4: Die bewegungskomponente. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/rob.beweg.ps.Z.
[ bib ]
[371] R. Der. Vorlesung Robotik. Universität Leipzig Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/skrpt-rob.html.
[ bib ]
[372] R. Der. Vorlesung Robotik. Kapitel 3. Universität Leipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/rob-neuro.ps.Z.
[ bib ]
[373] R. Der. Vorlesung Robotik. Kapitel 4: Roboterevolution. Universität Leipzig, Institut für Informatik.
[ bib ]
[374] R. Der. Vorlesung Robotik. Kapitel 5: Selbstorganisation und Artificial Life. Animaten. Universität Leipzig, Institut für Informatik.
[ bib ]
[375] R. Der. Vorlesung Robotik. Kapitel 6. Universität Leipzig, Institut für Informatik.
[ bib ]
[376] R. Der. Vorlesung Robotik. Kapitel 7. Universität Leipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/rob-kinmatik.ps.Z.
[ bib ]
[377] R. Der. Vorlesung Robotik. Kapitel 8. Universität Leipzig, Institut für Informatik. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/rob-dynamik.ps.Z.
[ bib ]
[378] R. Der. Vorlesung Robotik. UniversitätLeipzig Institut für Informatik kap. 2.3: Die planungskomponente. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/rob-plan.ps.Z.
[ bib ]
[379] R. Der. Vorlesung Robotik. UniversitätLeipzig Institut für Informatik kap. 2.1: Die sensorkomponente. http://www.informatik.uni-leipzig.de/~der/Vorlesungen/rob-sensor.ps.Z.
[ bib ]
[380] R. Der. Vorlesung Robotik.UniversitätLeipzig Institut für Informatik. kap. 2.2: Die wissenskomponente. http://www.informatik.uni-leipzig.de/{~}der/Vorlesungen/rob.wissen.ps.Z.
[ bib ]

Unsere Videos

[1] R. Der, F. Hesse, and G. Martius. Videos of self-organized creatures. http://robot.informatik.uni-leipzig.de/Videos, 2005.
[ bib ]
[2] R. Der, F. Hesse, and G. Martius. Video to the experiments section in this paper. http://robot.informatik.uni-leipzig.de/Videos/HurlingSnake/2005/CandA+.mpg, 2005.
[ bib ]
[3] R. Der, F. Hesse, and G. Martius. Video of spherical robot in a circular corridor. http://robot.informatik.uni-leipzig.de/Videos/SphericalRobot/2005/spherical_circular_corridor.mpg, 2005.
[ bib ]
[4] R. Der, F. Hesse, and G. Martius. Video of spherical robot in free space. http://robot.informatik.uni-leipzig.de/Videos/SphericalRobot/2005/spherical_floor_trace.mpg, 2005.
[ bib ]
[5] R. Der. Videos of self-organised robot behavior. http://robot.informatik.uni-leipzig.de/Videos, 2005.
[ bib ]
[6] R. Der. Video of robot in cluttered environment. http://robot.informatik.uni-leipzig.de/Videos/Pioneer/2004/maze2.wmv, 2004.
[ bib ]
[7] R. Der. Video of robot in cluttered environment. http://robot.informatik.uni-leipzig.de/Videos/Pioneer/2004/maze4.wmv, 2004.
[ bib ]
[8] R. Der. Basic homeokinetic control. http://robot.informatik.uni-leipzig.de/~der/Forschung//maze1.wmv, 2003.
[ bib ]
[9] R. Der. Basic homeokinetic control II. http://robot.informatik.uni-leipzig.de/~der/Forschung//maze2.wmv, 2003.
[ bib ]
[10] R. Der. Basic homeokinetic control III. http://robot.informatik.uni-leipzig.de/~der/Forschung//maze3.wmv, 2003.
[ bib ]
[11] R. Der. Basic homeokinetic control IV. http://robot.informatik.uni-leipzig.de/~der/Forschung//maze4.wmv, 2003.
[ bib ]

Ralf's Publikationen (generiert aus der.bib)

[1] F. Hesse, G. Martius, R. Der, and J. M. Herrmann. A sensor-based learning algorithm for the self-organization of robot behavior. Algorithms, 2(1):398-409, 2009. [ bib | http ]
Ideally, sensory information forms the only source of information to a robot. We consider an algorithm for the self-organization of a controller. At short time scales the controller is merely reactive but the parameter dynamics and the acquisition of knowledge by an internal model lead to seemingly purposeful behavior on longer time scales. As a paradigmatic example, we study the simulation of an underactuated snake-like robot. By interacting with the real physical system formed by the robotic hardware and the environment, the controller achieves a sensitive and body-specific actuation of the robot.

[2] F. Hesse, R. Der, and J. M. Herrmann. Reflexes from self-organizing control in autonomous robots. In L. Berthouze, C. G. Prince, M. Littman, H. Kozima, and C. Balkenius, editors, 7th International Conference on Epigenetic Robotics: Modelling Cognitive Development in Robotic Systems, Rutgers University, Piscataway, NJ, USA, volume 134 of Cognitive Studies, pages 37-44. Lund University, 2007. [ bib | .pdf ]
Homeokinetic learning provides a route to the self-organization of elementary behaviors in autonomous robots by establishing low-level sensomotoric loops. Strength and duration of the internal parameter changes which are caused by the homeokinetic adaptation provide a natural evaluation of external states, which can be used to incorporate information from additional sensory inputs and to extend the function of the low-level behavior to more general situations. We illustrate the approach by two examples, a mobile robot and a human-like hand which are driven by the same low-level scheme, but use the second-order information in different ways to achieve either risk avoidance and unconstrained movement or constrained movement. While the low-level adaptation follows a set of rigid learning rules, the second-order learning exerts a modulatory effect to the elementary behaviors and to the distribution of their inputs.

[3] G. Martius, J. M. Herrmann, and R. Der. Guided self-organisation for autonomous robot development. In A. e Costa and Francesco, editors, Advances in Artificial Life 9th European Conference, ECAL 2007, Lisbon, Portugal, volume 4648 of Lecture Notes in Computer Science, pages 766-775. Springer, 2007. [ bib | .pdf ]
The paper presents a method to guide the self-organised development of behaviours of autonomous robots. In earlier publications we demonstrated how to use the homeokinesis principle and dynamical systems theory to obtain self-organised playful but goal-free behaviour. Now we extend this framework by reinforcement signals. We validate the mechanisms with two experiment with a spherical robot. The first experiment aims at fast motion, where the robot reaches on average about twice the speed of a not reinforcement robot. In the second experiment spinning motion is rewarded and we demonstrate that the robot successfully develops pirouettes and curved motion which only rarely occur among the natural behaviours of the robot.

[4] R. Der, G. Martius, and F. Hesse. Let it roll - emerging sensorimotor coordination in a spherical robot. In L. M. Rocha, L. S. Yaeger, M. A. Bedau, D. Floreano, R. L. Goldstone, and A. Vespignani, editors, Artificial Life X : Proceedings of the Tenth International Conference on the Simulation and Synthesis of Living Systems, pages 192-198. International Society for Artificial Life, MIT Press, August 2006. [ bib | .pdf ]
Self-organization and the phenomenen of emergence play an essential role in living systems and form a challenge to artificial life systems. This is not only because systems become more life like but also since self-organization may help in reducing the design efforts in creating complex behavior systems. The present paper exemplifies a general approach to the self-organization of behavior which has been developed and tested in various examples in recent years. We apply this approach to a spherical robot driven by shifting internal masses. The complex physics of this robotic object is completely unknown to the controller. Nevertheless after a short time the robot develops systematic rolling movements covering large distances with high velocity. In a hilly landscape it is capable of manoeuvering out of the basins and in landscapes with a fixed rotational geometry the robot more or less adatps its movements to this geometry - the controller so to say develops a kind of feeling for its environment although there are no sensors for measuring the positions or the velocity of the robot. We argue that this behavior is a result of the spontaneous symmetry breaking effects which are responsible for the emergence of behavior in our approach.

[5] R. Der and G. Martius. From motor babbling to purposive actions: Emerging self-exploration in a dynamical systems approach to early robot development. In S. Nolfi, G. Baldassarre, R. Calabretta, J. C. T. Hallam, D. Marocco, J.-A. Meyer, O. Miglino, and D. Parisi, editors, From Animals to Animats 9, 9th International Conference on Simulation of Adaptive Behavior, SAB 2006, Rome, Italy, September 25-29, 2006, Proceedings, volume 4095 of Lecture Notes in Computer Science, pages 406-421. Springer, 2006. [ bib | .pdf ]
Self-organization and the phenomenon of emergence play an essential role in living systems and form a challenge to artificial life systems. This is not only because systems become more lifelike, but also since self-organization may help in reducing the design efforts in creating complex behavior systems. The present paper studies self-exploration based on a general approach to the self-organization of behavior, which has been developed and tested in various examples in recent years. This is a step towards autonomous early robot development. We consider agents under the close sensorimotor coupling paradigm with a certain cognitive ability realized by an internal forward model. Starting from tabula rasa initial conditions we overcome the bootstrapping problem and show emerging self-exploration. Apart from that, we analyze the effect of limited actions, which lead to deprivation of the world model. We show that our paradigm explicitly avoids this by producing purposive actions in a natural way. Examples are given using a simulated simple wheeled robot and a spherical robot driven by shifting internal masses.

[6] R. Der, F. Hesse, and G. Martius. Rocking stamper and jumping snake from a dynamical system approach to artificial life. Adaptive Behavior, 14(2):105-115, 2006. [ bib | DOI | .pdf ]
Keywords: autonomous robots, self-organization, homeostasis, homeokinesis, dynamical systems, learning
[7] R. Der. Homeokinesis and the moderation of complexity in neural systems. Neural Computation, to be submitted, 2005. [ bib | .pdf ]
[8] R. Der, F. Hesse, and G. Martius. Learning to feel the physics of a body. In CIMCA '05: Proceedings of the International Conference on Computational Intelligence for Modelling, Control and Automation and International Conference on Intelligent Agents, Web Technologies and Internet Commerce Vol-2 (CIMCA-IAWTIC'06), pages 252-257, Washington, DC, USA, 2005. IEEE Computer Society. [ bib | .pdf ]
Despite the tremendous progress in robotic hardware and in both sensorial and computing efficiencies the performance of contemporary autonomous robots is still far below that of simple animals. This has triggered an intensive search for alternative approaches to the control of robots. The present paper exemplifies a general approach to the self-organization of behavior which has been developed and tested in various examples in recent years. We apply this approach to an underactuated snake like artifact with a complex physical behavior which is not known to the controller. Due to the weak forces available, the controller so to say has to develop a kind of feeling for the body which is seen to emerge from our approach in a natural way with meandering and rotational collective modes being observed in computer simulation experiments.

[9] R. Der, F. Hesse, and R. Liebscher. Contingent robot behavior generated by self-referential dynamical systems. Autonomous robots, 2005. submitted. [ bib | .pdf ]
[10] R. Der. Videos of self-organised robot behavior. http://robot.informatik.uni-leipzig.de/Videos, 2005. [ bib ]
[11] R. Der. Between homeostasis and autopoiesis - theory and practice of self-referential machines. Talk given at the Max-Planck Institut for Mathematics in the Sciences, October 2004. [ bib ]
[12] R. Der. The homeokinetic neuron in the sensorimotor loop. Neural computation, 2004. to be submitted. [ bib | .pdf ]
[13] M. Herrmann, M. Hoicki, and R. Der. On ashby's homeostat: A formal model of adaptive regulation. In S. Schaal, editor, From Animals to Animats, pages 324 - 333. MIT Press, 2004. [ bib | .pdf ]
[14] R. Der, F. Hesse, and R. Liebscher. Self-organized exploration and automatic sensor integration from the homeokinetic principle. In Proc. 3rd Workshop on Self-Organization of AdaptiVE Behavior (SOAVE'04), Fortschritt-Berichte VDI, Reihe 10, Nr. 743, pages 220-230. VDI-Verlag, 2004. [ bib | .pdf ]
[15] R. Der. Video of robot in cluttered environment. http://robot.informatik.uni-leipzig.de/Videos/Pioneer/2004/maze2.wmv, 2004. [ bib ]
[16] R. Der. Video of robot in cluttered environment. http://robot.informatik.uni-leipzig.de/Videos/Pioneer/2004/maze4.wmv, 2004. [ bib ]
[17] R. Der. Basic homeokinetic control. http://www.informatik.uni-leipzig.de/~der/Forschung/maze1.wmv, 2003. [ bib ]
[18] R. Der. Basic homeokinetic control II. http://www.informatik.uni-leipzig.de/~der/Forschung/maze2.wmv, 2003. [ bib ]
[19] R. Der. Basic homeokinetic control III. http://www.informatik.uni-leipzig.de/~der/Forschung/maze3.wmv, 2003. [ bib ]
[20] R. Der. Basic homeokinetic control IV. http://www.informatik.uni-leipzig.de/~der/Forschung/maze4.wmv, 2003. [ bib ]
[21] R. Der, M. Herrmann, and R. Liebscher. Homeokinetic approach to autonomous learning in mobile robots. In R. Dillman, R. D. Schraft, and H. Wörn, editors, Robotik 2002, number 1679 in VDI-Berichte, pages 301-306. 2002. [ bib ]
[22] R. Der, M. Herrmann, and M. Holicki. Self-organization in sensor-motor loops by the homeokinetic principle. Verhandlungen der Deutschen Physikalischen Gesellschaft, page 510, 1 2002. [ bib ]
[23] R. Der and R. Liebscher. True autonomy from self-organized adaptivity. In Proc. Workshop Biologically Inspired Robotics, Bristol, 2002. [ bib | .pdf ]
[24] R. Der. Self-organized acquisition of situated behavior. Theory Biosci., 120:179-187, 2001. [ bib ]
[25] R. Der. Self-organized robot behavior from the principle of homeokinesis., 1999. [ bib | .ps ]
[26] R. Der, U. Steinmetz, and F. Pasemann. Homeokinesis - a new principle to back up evolution with learning. In Computational Intelligence for Modelling, Control, and Automation, volume 55 of Concurrent Systems Engineering Series, pages 43-47, Amsterdam, 1999. IOS Press. [ bib | .ps ]
[27] R. Der, O. Lummer, and T. List. Incremental nonlinear dynamic data reconciliation. Technical Report 3/98, Institut für Informatik, Universität Leipzig, 1997. [ bib | .ps ]
[28] R. Der and M. Herrmann. Self-adjusting reinforcement learning. In Nonlinear Theory and Applications - NOLTA 96, pages 441 - 444, 1996. [ bib | .ps.gz ]
[29] R. Der and M. Herrmann. Efficient Q-learning by division of labour. In Proc. International Conference on Artificial Neural Networks - ICANN95, pages 129 - 134, 1995. [ bib | .ps.gz ]
[30] R. Der. The langevin method in the dynamics of learning. J. Phys. A: Math. Gen., (23):L763-6, 1990. [ bib ]
[31] R. Der. Systems under colored noise. Physica A, 154:421 - 451, 1989. [ bib ]
[32] R. Der. Systems under colored noise. Physica A, 154:421-451, 1989. [ bib ]
[33] R. Der. The time local view of nonequilibrium statistical mechanics I. J. Stat. Phys., 46:349-390, 1987. [ bib ]
[34] R. Der. The time local view of nonequilibrium statistical mechanics II. J. Stat. Phys., 46:391-425, 1987. [ bib ]
[35] R. Der. The time local view of nonequilibrium statistical mechanics I. J. Stat. Phys., 46:349-390, 1987. [ bib ]
[36] R. Der. The time local view of nonequilibrium statistical mechanics II. J. Stat. Phys., 46:391-425, 1987. [ bib ]
[37] R. Der. Retarded and instantaneous evolution equations for macroobsevables in nonequilibrium statistical mechanics. Physica, 132A:74-93, 1985. [ bib ]