Web Traffic Forecasting Using ARIMA and LSTM

In today’s world web traffic is one of the serious issues faced by many. Web traffic tends to hinder the smooth user experience and it is also very challenging for the web service providers to maintain a smooth user-server interaction. We are looking to overcome this problem by building a prediction model to forecast the web traffic in advance to avoid all the problems faced. Our model thoroughly studies the previous web traffic data to efficiently predict the web traffic of a particular website at a given point in time. Forecasting is one of the important goals of mining time-series databases. The efficacy of Time series forecasting has been proved while decision making in various domains. This method is vastly different from the other proposed methods for prediction and analysis. This paper proposes the use of ARIMA and LSTM algorithms to forecast web traffic.