ОСОБЛИВОСТІ СУЧАСНИХ МЕТОДІВ МАШИННОГО НАВЧАННЯ ЩОДО ВИРІШЕННЯ ЗАДАЧІ ПРОГНОЗУВАННЯ ДАНИХ

This work is devoted to the process of analyzing and identifying the features of machine learning methods in solving data prediction problems. The relevance of the forecasting task in the modern world is described, and the importance of choosing the right method is mentioned. The following tools were considered: quantile regression forests, recurrent neural networks and support vector method. The areas in which the tools listed are commonly used are described. Problems that can be encountered when working with it. Some recommendations are provided that may help to improve the quality of the models.