Learning automata based method for solving demand and supply problem with periodic behaviors

Haoye Lu, Anand Srinivasan, Amiya Nayak

Abstract: The demand and supply problems are encountered widely in almost all fields. While the users hope to get their demand satisfied always, the suppliers do not want to produce superfluous supplies. One way to tackle this problem is to have a good prediction on the future demand. Many forecasting algorithms have been proposed by using time series and extrapolation techniques. However, most of the algorithms are based on the complex theories and consume potentially large amount computing resources. In this paper, we propose a new algorithm to predict the demand by applying classical learning automata techniques. The algorithm uses more concise strategies and consumes considerably less computing resources.