Adaptive Multiregression Currency Estimation in the Chaotic Market Environment

The problem of multiregression estimation of the currency cost is considered. The offered approach is based on an adaptive choice of the regressors formed by group of currency pairs, the most correlated with an estimated asset. In the conditions of chaotic dynamics of currency quotations, correlation degree between currency pairs changes in time. From here the problem of adaptive estimation with variable structure of group of regressors follows. The method of evolutionary modeling is used for an assessment of the potential prize, reached when using the corresponding control strategy.