Using a new type of nonlinear integral for multi-regression: An application of evolutionary algorithms in data mining

Abstract

A new nonlinear multiregression model, Y = c+q·(w)∂f dμ+N(0,σ2), based on the Wang integral is developed to describe the multi-input single-output system. The model includes a non-additive set function μ, to describe the inherent interaction among the input attributes x1, x2, ..., xn. When the proper input-output data are available, by using the adaptive genetic algorithm, a precise estimated values of parameter c, q, ω, and μ of the regression model can be obtained. The output, Y, could be calculated by nonlinear multiregression using the values of known input attributes, x1, x2, ..., xn.