The implementation of predictive graph extrapolation method

Abstract

This project will implement the predictive method for graph extrapolation in time series data. The project compared the performance between predictive method of Kalman filter and Holt-Winters with estimating technique. The project also shows that the estimating technique is not realiableto forecast future value and may lead to misleading decision. The predictive method is a better option for forecasting data which can lead to better desicion making. The predictive method has smaller error compared to estimating technique.