Machine learning, Prophet and XGBoost algorithm: Analysis of Traffic Forecasting in Telecom Networks with time series data
Garima Jain, Rajeev Ranjan Prasad
Abstract
Network traffic or data traffic is defined as the volume of data moving across a network at a given point of time. Network data in computer networks is usually encapsulated in system packets, which describe the load within the network. Network traffic is the main component for the traffic measurement, traffic control and simulation. The forecasting of each indexes of telecom services for the business planning purpose has become more and more important, especially for telecom operators. The forecasting results will help to create an optimal business planning for the purpose of capacities, investments and resources. The research in the field of network traffic prediction using machine learning has been reported in literatures. However, those researches are inadequate and provides insufficient knowledge about the issues like tuning of time-series data, modelling and even the acceptance requirement from customer. This paper focuses on these issues of traffic forecasting of telecom networks with time series data with the help of machine learning, prophet algorithm and XGBoost algorithm.