Author:

Keywords:

rbf networks, gaussian mixture distribution, global optimization, fokker-planck equation, constrained lms, regularization, stochastic-approximation, neural networks, algorithms, behavior, rd, SISTA, Science & Technology, Technology, Computer Science, Artificial Intelligence, Computer Science, RBF networks, Gaussian mixture distribution, Fokker-Planck equation, constrained LMS, STOCHASTIC-APPROXIMATION, GLOBAL OPTIMIZATION, NEURAL NETWORKS, ALGORITHMS, BEHAVIOR, RD, 0801 Artificial Intelligence and Image Processing, 1702 Cognitive Sciences, Artificial Intelligence & Image Processing, 4602 Artificial intelligence, 4611 Machine learning

Abstract:

In this letter we present an on-line learning version of the Fokker-Planck machine. The method makes use of a regularized constrained normalized LMS algorithm in order to estimate the time-derivative of the parameter vector of a radial basis function network. The RBF network parametrizes a transition density which satisfies a Fokker-Planck equation, associated to continuous simulated annealing. On-line learning using the constrained normalized LMS method is necessary in order to make the Fokker-Planck machine applicable to large scale nonlinear optimization problems.