Comparative analysis of methods of forecasting the consumer price index for food products (on the example of the Altai territory)

Abstract

At the moment, there are no uniform universal methods for forecasting regional indicators of economic development in general and the consumer price index in particular. But depending on how accurate and reasonable the forecasts of the consumer price index will be, the budget of the region will be drawn up so correctly and the parameters of the forecast of socio-economic development, in the calculation of which this indicator is used, will be accurately predicted. The article presents a comparative analysis of methods for forecasting the consumer price index for food products. First, the most popular methods of forecasting the consumer price index were identified. Then models of time series, neural networks and decision trees were built, as well as retro-forecasts of the consumer price index for food products based on them. It is revealed that neural networks provide higher accuracy of forecasts compared to other models. The result of the work was the forecast of the consumer price index for food products in the Altai Territory for 2021 based on the constructed neural network model. The constructed neural network models can be used in relevant organizations to increase the accuracy of the forecast of this indicator. In addition, such an approach can be used as a basis for forecasting other indicators that characterize the socio-economic development of regions.