Life-long learning Cell Structures--continuously learning without catastrophic interference

As an extension of on-line learning, life-long learning challenges a system which is exposed to patterns from a changing environment during its entire lifespan. An autonomous system should not only integrate new knowledge on-line into its memory, but also preserve the knowledge learned by previous interactions. Thus, life-long learning implies the fundamental Stability-Plasticity Dilemma, which addresses the problem of learning new patterns without forgetting old prototype patterns. We propose an extension to the known Cell Structures, growing Radial Basis Function-like networks, that enables them to learn their number of nodes needed to solve a current task and to dynamically adapt the learning rate of each node separately. As shown in several simulations, the resulting Life-long Learning Cell Structures posses the major characteristics needed to cope with the Stability-Plasticity Dilemma.

[1]  Michael McCloskey,et al.  Catastrophic Interference in Connectionist Networks: The Sequential Learning Problem , 1989 .

[2]  Dietmar Heinke,et al.  Comparing neural networks: a benchmark on growing neural gas, growing cell structures, and fuzzy ARTMAP , 1998, IEEE Trans. Neural Networks.

[3]  Bernd Fritzke,et al.  Growing cell structures--A self-organizing network for unsupervised and supervised learning , 1994, Neural Networks.

[4]  GrossbergS..  Adaptive pattern classification and universal recoding , 1976 .

[5]  Elie Bienenstock,et al.  Neural Networks and the Bias/Variance Dilemma , 1992, Neural Computation.

[6]  Mark S. Nixon,et al.  Generating-shrinking algorithm for learning arbitrary classification , 1994, Neural Networks.

[7]  Thomas Martinetz,et al.  Topology representing networks , 1994, Neural Networks.

[8]  D. Signorini,et al.  Neural networks , 1995, The Lancet.

[9]  Nicolaos B. Karayiannis,et al.  Reformulated radial basis neural networks trained by gradient descent , 1999, IEEE Trans. Neural Networks.

[10]  David Saad,et al.  Online Learning in Radial Basis Function Networks , 1997, Neural Computation.

[11]  Hilbert J. Kappen,et al.  On-line learning processes in artificial neural networks , 1993 .

[12]  Horst-Michael Gross,et al.  Sensory-based Robot Navigation using Self-organizing Networks and Q-learning , 1996 .

[13]  Fred Henrik Hamker.  Visuelle Aufmerksamkeit und lebenslanges Lernen im Wahrnehmungs-Handlungs-Zyklus , 1999, Künstliche Intell..

[14]  Thomas Martinetz,et al.  'Neural-gas' network for vector quantization and its application to time-series prediction , 1993, IEEE Trans. Neural Networks.

[15]  T. Heskes,et al.  On-line Learning Processes in Artiicial Neural Networks On-line Learning Processes in Artiicial Neural Networks , 2007 .

[16]  Christian Lebiere,et al.  The Cascade-Correlation Learning Architecture , 1989, NIPS.

[17]  T Poggio,et al.  Regularization Algorithms for Learning That Are Equivalent to Multilayer Networks , 1990, Science.

[18]  Mahesan Niranjan,et al.  A dynamic neural network architecture by sequential partitioning of the input space , 1993, IEEE International Conference on Neural Networks.

[19]  Andreas Ziehe,et al.  Adaptive On-line Learning in Changing Environments , 1996, NIPS.

[20]  Haim Sompolinsky,et al.  On-line Learning of Dichotomies: Algorithms and Learning Curves. , 1995, NIPS 1995.

[21]  S. Amari.  A Theory ofAdaptive Pattern Classifiers , 1967 .

[22]  Bernd Fritzke,et al.  A Growing Neural Gas Network Learns Topologies , 1994, NIPS.

[23]  Bernd Fritzke,et al.  A Self-Organizing Network that Can Follow Non-stationary Distributions , 1997, ICANN.

[24]  D. Saad,et al.  Dynamics of on-line learning in radial basis function networks , 1997 .

[25]  Stephen Grossberg,et al.  Fuzzy ARTMAP: A neural network architecture for incremental supervised learning of analog multidimensional maps , 1992, IEEE Trans. Neural Networks.

[26]  CHEE PENG LIM,et al.  An Incremental Adaptive Network for On-line Supervised Learning and Probability Estimation , 1997, Neural Networks.

[27]  Stephen Grossberg,et al.  ARTMAP: supervised real-time learning and classification of nonstationary data by a self-organizing neural network , 1991, [1991 Proceedings] IEEE Conference on Neural Networks for Ocean Engineering.

[28]  Gerald Sommer,et al.  Dynamic Cell Structure Learns Perfectly Topology Preserving Map , 1995, Neural Computation.

[29]  R. French.  Catastrophic forgetting in connectionist networks , 1999, Trends in Cognitive Sciences.

[30]  Jooyoung Park,et al.  Approximation and Radial-Basis-Function Networks , 1993, Neural Computation.

[31]  Dragan Obradovic,et al.  On-line training of recurrent neural networks with continuous topology adaptation , 1996, IEEE Trans. Neural Networks.

[32]  D. Broomhead,et al.  Radial Basis Functions, Multi-Variable Functional Interpolation and Adaptive Networks , 1988 .

[33]  Stephen Grossberg,et al.  Nonlinear neural networks: Principles, mechanisms, and architectures , 1988, Neural Networks.

[34]  Michael R. Berthold,et al.  Boosting the Performance of RBF Networks with Dynamic Decay Adjustment , 1994, NIPS.

[35]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[36]  Bruce A. Whitehead,et al.  Evolving space-filling curves to distribute radial basis functions over an input space , 1994, IEEE Trans. Neural Networks.

[37]  Michael Hansen,et al.  Adaptive Saccade Control of a Binocular Head with Dynamic Cell Structures , 1996, ICANN.

[38]  Gerald Sommer,et al.  An integrated architecture for learning of reactive behaviors based on dynamic cell structures , 1997, Robotics Auton. Syst..

[39]  John C. Platt.  A Resource-Allocating Network for Function Interpolation , 1991, Neural Computation.

[40]  F. Gage,et al.  Neurogenesis in the adult human hippocampus , 1998, Nature Medicine.

[41]  S. Grossberg,et al.  The Hippocampus and Cerebellum in Adaptively Timed Learning, Recognition, and Movement , 1996, Journal of Cognitive Neuroscience.

[42]  Paramasivan Saratchandran,et al.  Performance evaluation of a sequential minimal radial basis function (RBF) neural network learning algorithm , 1998, IEEE Trans. Neural Networks.

[43]  T. Kohonen.  Self-organized formation of topographically correct feature maps , 1982 .

[44]  Jack F. Gerrissen,et al.  On the network-based emulation of human visual search , 1991, Neural Networks.

[45]  Asim Roy,et al.  A neural-network learning theory and a polynomial time RBF algorithm , 1997, IEEE Trans. Neural Networks.

[46]  Nicolaos B. Karayiannis,et al.  Growing radial basis neural networks: merging supervised and unsupervised learning with network growth techniques , 1997, IEEE Trans. Neural Networks.

[47]  R Ratcliff,et al.  Connectionist models of recognition memory: constraints imposed by learning and forgetting functions. , 1990, Psychological review.

[48]  Donald F. Specht,et al.  Probabilistic neural networks , 1990, Neural Networks.

[49]  James L. McClelland,et al.  Why there are complementary learning systems in the hippocampus and neocortex: insights from the successes and failures of connectionist models of learning and memory. , 1995, Psychological review.

[50]  David S. Broomhead,et al.  Multivariable Functional Interpolation and Adaptive Networks , 1988, Complex Syst..

[51]  Jörg Bruske.  Dynamische Zellstrukturen: Theorie und Anwendung eines KNN-Modells , 1998, Ausgezeichnete Informatikdissertationen.