Adaptive modeling and forecasting of seasonal phenomena based on a multiplicative model

Abstract

Importance: due to the natural processes of growth of the world population and temperature on the planet there is a problem of water shortage. Now on the stock exchange, the water resource is on a par with oil, gas, gold and so on. The emergence of water futures blew up the world community, removing such an urgent problem from the category of not only environmental, but also economic. Purpose: in this regard, there is an urgent need to select adequate tools for forecasting water demand, which is associated with climate and periodicity, and as a result of this, as a mathematical toolkit, the authors proposed adaptive models of seasonal phenomena, and as an object of research – time series of monthly sales of drinking water. Research design: regulation of an unevenly distributed vital resource requires in the current situation the involvement of the economic and mathematical apparatus in order to promptly solve the issues of modeling and forecasting the behavior of the water market as a whole and its individual components. In connection with the indicated problem, the authors considered the prognostic potential of the adaptive model as an effective toolkit for short-term forecasting the volume of water consumption in PET bottles in the Krasnodar Territory. Results: the result of the study is the construction of an adaptive model that demonstrates the minimum values of the maximum relative error of the forecast and allows on its basis to obtain a close to real extrapolation of the dynamics of water consumption in the studied time series.