Title :
Unsupervised learning in noise
Author_Institution :
Dept. of Electr. Eng., Univ. of Southern California, Los Angeles, CA, USA
fDate :
3/1/1990 12:00:00 AM
Abstract :
A new hybrid learning law, the differential competitive law, which uses the neuronal signal velocity as a local unsupervised reinforcement mechanism, is introduced, and its coding and stability behavior in feedforward and feedback networks is examined. This analysis is facilitated by the recent Gluck-Parker pulse-coding interpretation of signal functions in differential Hebbian learning systems. The second-order behavior of RABAM (random adaptive bidirectional associative memory) Brownian-diffusion systems is summarized by the RABAM noise suppression theorem: the mean-squared activation and synaptic velocities decrease exponentially quickly to their lower bounds, the instantaneous noise variances driving the system. This result is extended to the RABAM annealing model, which provides a unified framework from which to analyze Geman-Hwang combinatorial optimization dynamical systems and continuous Boltzmann machine learning
Keywords :
content-addressable storage; encoding; learning systems; neural nets; noise; Boltzmann machine learning; Brownian-diffusion systems; Gluck-Parker; Hebbian learning systems; RABAM; coding; differential competitive law; neural nets; neuronal signal velocity; noise suppression; random adaptive bidirectional associative memory; unsupervised learning; Associative memory; Biological system modeling; Large-scale systems; Neural networks; Neurofeedback; Stability; Stochastic resonance; Stochastic systems; Structural engineering; Unsupervised learning;
Journal_Title :
Neural Networks, IEEE Transactions on
