The Investigation of Machine Learning Methods in the Problem of Automation of the Sales Management Business-process

Methods of machine learning in the application to the task of managing sales in retail trade are considered. The features of the sales management process and the nuances of applying the methods of machine learning are noted. The advantages of using machine learning methods are highlighted. Attention is focused on the risks in the application of methods of machine learning. Practical methods of machine learning "without a teacher" are considered in practice. The quality of the results of the applied methods is estimated using various quality metrics. As a result, it was found that reducing the dimension of the feature space does not lead to loss of information, but helps reduce the load on the computing system.