An Experimental Study for Comparing Different Method for Time Series Forecasting Prediction & Anomaly Detection

Abstract

Time series forecasting is used to detect some anomaly, that is, any unusual or unrequired events in network traffic, so that it can be removed while using the dataset for further processing. Anomaly detection is very helpful in reducing the operation call. This paper compares different models for detecting anomaly in computer networks using time series forecasting methods with reduced error rates.