An Integrated Forecasting Technique with Modified Weight Measurement

Abstract

Forecasting has long been part of our life since early of the history of human being. In the middle of 20th century forecasting becomes a part of every business and financial sectors. Nowadays every successful firm has to make its own forecasts with an acceptable error as there is no chance of zero error. The situation becomes more complicated if the observed data is more diverted from the existing pattern. In such situation it becomes more difficult to fit it into a suitable forecasting model.Then it requires to combine several forecasts to reach a better forecast.In this paper, we willdevelop a sophisticated forecasting technique bycombining the weighted average method with Linear Programming (LP) model by developing an alternative technique to calculate the weights. We will carry out our analysis by using Microsoft Excel, statistical data analysis tool R and MATHEMATICA. We will demonstrate our model by numerical examples. Dhaka Univ. J. Sci. 71(1): 36-41, 2023 (Jan)