Bioproductivity of Dnieper Reservoirs Analysis by Inductive Methods

Abstract

The task was to determine the factors that have the most significant influence on the state of water in the Dnieper reservoirs by constructing a model of dependence of the concentration of chlorophyll a in phytoplankton according to long-term observations in Kremenchug and Kakhovka reservoirs. The results of observations of the Institute of Sciences in 1976 – 1993. Various inductive methods were used to obtain a satisfactory result. Algorithms: linear regression of (LR), LASSO, combinatorial algorithm of (COMBI) GMDH and correlation-rating algorithm (CRA).