MODELS AND METHODS FOR FORECASTING FINANCIAL TIME SERIES ON THE EXAMPLE OF THE USD / RUB COURSE

UDC 336.76

Authors

V.A. IVANYUK, candidate of economic sciences, associate professor

Abstract

The article provides a classification of forecasting methods. The analysis of forecasting models is carried out. An aggregate forecasting technique is proposed based on three forecasting methods: linear forecasting, asymptotic forecasting and neural forecasting. Based on the financial time series of USD / RUB, an aggregate forecast is built and its accuracy is calculated.

Keywords

forecasting models, time series, trend, neural forecast.