Dynamic trustee Behaviour pattern for user support and its forecast prediction

Abstract

The Major focuses on the various trustee behavior patterns namely random, stable, trend, jumping and two phases with data analysis. The Forecast was based on the predictions between the previous predictions of mean square error and its ratings were discussed. At that point figure was made for the reasonable model with the moving normal, weighted moving normal, single exponential smoothing, twofold exponential smoothing triple exponential smoothing and holt's direct changes. The value of double exponential smoothing with the five behavior patterns yields larger mean square errors which had been further quantified with data analysis. The performance metrics of each had been made with forecast and its moving average error.