Interacting neural networks

Several scenarios of interacting neural networks which are trained either i n an identical or in a competitive way are solved analytically. In the case of identical training each perceptron receives the output of its neighbor. The symmetry of the stationary state as well as the sensitivity to the use d training algorithm are investigated. Two competitive perceptrons trained on mutually exclusive learning aims and a perceptron which is trained on th e opposite of its own output are examined analytically. An ensemble of comp etitive perceptrons is used as decision-making algorithms in a model of a c losed market (El Farol Bar problem or the Minority Game. In this game, a se t of agents who have to make a binary decision is considered.); each networ k is trained on the history of minority decisions. This ensemble of percept rons relaxes to a stationary state whose performance can be better than ran dom.