Scholarly record

PLANNING BUSINESS EXPENSES THROUGH SHORT-TERM FORECASTING

Krasimira Stoilova, Todor Stoilov

Abstract

The research is aimed at developing intelligent solution for the manager to better plan future resources. A key element in profit management is resource allocation. The idea of this research is based on the desire to improve the profit of a business organization. Quantitative formalization of this process is done by presenting mathematical models for predicting. Formal modeling in forecasting is based on statistical analysis of available historical data and estimation of future behavior. Mathematical prediction models most often use linear regression as a methodology. In the present study, the first-order and second-order AR formalization is used and a comparison is made between them in terms of prediction accuracy. We use real data on electricity costs on a business entity. We want to use this data to predict future costs by applying the two AR(1) and AR(2) linear regression models. We define a sliding prediction policy with different values of the history interval. It ranges from 4 to 8 to analyze the influence of the historical interval on the modeling accuracy. The corresponding values of the regression coefficients and predicted values are mathematically determined. The predicted values are compared with the actual values. The two types of predicted values from the two linear auto-regression models are compared and analyzed. The comparison between the AR(1) and AR(2) models gives a preference in terms of the accuracy of the AR(2) model. The forecasted values are useful for the manager to better plan the required funds and for better market positioning.

Publication details

Back to publication list