Development and Research of the Forecasting Models Based on the Time Series Using the Random Forest Algorithm

The problem of developing forecasting models based on multidimensional time series, which are a kind of features and are used in the formation of the datasets, dividing further into the training and test sets, using the example of the problem of forecasting the remaining useful life of the complex technical systems, has been considered. The possibilities of the random forest algorithm when constructing the regression model have been investigated. The approaches to improving the forecasting accuracy based on the regression model, which consist in generating new features to be included in the training and test datasets have been considered. The recommendations for the development of the regression models for predicting the remaining useful life have been formed. The examples of forecasting the remaining useful life using the considered algorithms and approaches have been given.