Nonlinear Dynamics, Psychology, and Life Sciences, Vol. 10, Iss. 3, July, 2006, pp. 341-364
@2006 Society for Chaos Theory in Psychology & Life Sciences


Learning by Operant Conditioning as a Nonlinear Self-organized Process

M. A. Pascual, CAMP, Burgos, Spain
M. A. Rodríguez, IFCA, CSIC-Universidad de Cantabria, Santander, Spain

Abstract: Responses of individuals have been rewarded in fixed ratio operant conditioning experiments throughout the last 5 years. The long numerical series of performance rates so obtained have been analyzed by using Fourier spectra, obtaining 1/f noise, and wavelet techniques for the analysis of local variability. These techniques reveal that the evolution of data in each individual is nonlinear but it fluctuates showing self-organized patterns in successive time scales, which leads to a long term logarithmic growth. Based on these observations a model has been developed. It can simulate with a great accuracy the rate of response exhibited by our subjects in all time scales.

Keywords: operant learning, nonlinear processes, fractals, stochastic processes, 1/f Noise