Study of Supervised Algorithms for Solve the Forecasting Retail Dynamics Problem

In this article, we examine the use of machine learning models to predict retail dynamics. The main goal of this article is to analyze existing approaches and draw a conclusion about the advantages and disadvantages of each method. Main approaches and practical examples of using machine learning to solve the problem of forecasting retail dynamics are considered in practice. The advantages of using various machine learning methods, such as learnability, dynamism, nonlinearity and others are shown. The quality of the results of the applied methods is estimated using various quality metrics. A hierarchy of algorithms was proposed that showed better results than individual algorithms. This allows us to conclude that the application of the proposed hierarchy of algorithms improves the quality for predicting the dynamics of retail trade.