FUZZY COGNITIVE MAP VS REGRESSION

摘要

A fuzzy cognitive map is considered as an alternative to regression analysis, i.e., tools for modeling the inputs-output dependence based on expert-experimental information. To calculate the output value at the given input values, increments of variables are used. The optimal values of the weights of the arcs are found using the genetic algorithm in which the chromosomes are generated from the intervals of their feasible values and the selection criterion is the sum of the squared deviations between the model and the observed output values.