A classification method of network business type based on XGboost

In order to distinguish different types of network services and analyze the needs of customers of different types of services to ensure user perception. This paper proposes a classification method of network business type based on XGboost. Firstly, a feature contribution analysis method based on Pearson correlation coefficient is proposed, then the classification model of network business type based on XGboost is constructed to classify network business type. Finally, RMSE is used to evaluation the classification method. Compared with other models, the RMSE of XGboost model is no more than 3.5, lower than LSTM, ARIMA, LR and SVM model, which proves the reliability of the XGboost model in network business type classification.