The Comparative Evaluation of the Additive Mathematical Models and Tools for Time Series Analysis on the Example of Economic and Financial Indicators.

In this paper, we carry out a comparative description of time series forecasting models and analytics and business intelligence (ABI) systems that allow to implement these models. The study is devoted to the problem of day-trading, which requires predicting the sign of the difference between the next day price and the current price of an asset that is traded. In addition to well-known forecasting models, an algorithm is proposed in the work that implements trading based on the statistical characteristics of the price time series. Moreover, we solve the problem of forecasting a time series with long-term memory using the example of a series of values of the Gross Domestic Product. Both problems are related to the usage of exponential smoothing and ARIMA (p, d, q) models. The practical part was performed using the Microsoft Power BI, Tableau and SAS Studio platforms. The results of the comparison of the effectiveness of which are also presented in this paper.