Network Quality Operation Prediction Based on Machine Learning Algorithms

Abstract

The article proposes a solution to the problem of predicting the quality of a computer network using a popular machine learning method - a decision tree. Considered a real case of quality problems in the network when the web server is running. Based on the analysis of traffic recorded during the operation of the network, when problems were observed in it, a decision tree was constructed. Using the obtained model, numerical experiments were carried out to predict the moments when users have problems opening web pages. The results of assessing the quality of the prediction model obtained are given.