Daily Network Traffic Prediction Based on Backpropagation Neural Network

Abstract

Background: The analyzing and predicting network traffic usage is a very important issue in the service activities of the university. Objective: This paper presents the development of Backpropagation neural network (BPNN) algorithms for analyzing and predicting daily network traffic. Results: The gradient descent with momentum (traingdm) algorithm, and two-hidden layers (5-10-5-1) can be used as a model to predict the future. Conclusion: The BPNN technique has been able to approach the performance goals, and also has a pretty good MSE value.