Analysis of the Application of Machine Learning for the Tasks of Telecom Operators

Abstract

In the transition to fifth generation networks, telecom operators are faced with more and more tasks that require solutions using complex data processing algorithms, in particular – forecasting tasks. Machine learning is becoming increasingly popular for application, but due to the variety of its algorithms and the impossibility of their universal use, there is a need to determine the area of application for each of them. For telecom operators one of the priority tasks is to predict various parameters (contact center load, subscriber churn, loyalty index) in an enormous growth of the number of interacting objects, diversity of services, increasing data volumes. The authors of the report find it interesting to find the connection be-tween the tasks of telecoms operator and the methods of machine learning, which can be used to effec-tively solve these problems.