Estimating Time Series Trend Using Linear Programming Techniques On The Sales Data Of Kaduna Coca-Cola Bottling Company

Abstract

In this paper, we apply linear programming techniques to estimate trend of the sales data of Kaduna Coca-Cola Bottling Company. The trend component of any data has two important properties: smoothness and fidelity (closeness to the data). The linear programming (LP) solution is a monotone sequence which optimizes some weighted combination of both properties.