Abstract:
The aim of the chapter is to describe advanced methods of forecasting used in
modern, high-tech enterprises. One of them – the ARMA model considers the
strong dependence between the individual observations, used for prediction of
time series, characterized by high dynamics of change. Here will be explained
the process of selection of its parameters, and method design of the model.
Second – harmonic analysis – uses, in turn, the cyclicality of the time series
by which to construct the model describing the time series and what is the
forecast for future periods. The first method is a group of parametric models,
the second one is nonparametric, and both use the nature of a change of time
series. The use of these methods will be shown by example.