Title: Prediction of retail prices of products using local competitors

Authors: Hassan Waqar Ahmad; Sandra Zilles; Howard J. Hamilton; Richard Dosselmann

Addresses: Department of Computer Science, University of Regina, 3737 Wascana Parkway, Regina, SK S4S 0A2, Canada ' Department of Computer Science, University of Regina, 3737 Wascana Parkway, Regina, SK S4S 0A2, Canada ' Department of Computer Science, University of Regina, 3737 Wascana Parkway, Regina, SK S4S 0A2, Canada ' Department of Computer Science, University of Regina, 3737 Wascana Parkway, Regina, SK S4S 0A2, Canada

Abstract: Businesses and customers are interested in predicting the retail prices of products. In a competitive environment, the price of a product at a given target outlet is typically related to the price of the same or similar products at nearby competing outlets. This research predicts the start of day and current prices of a specific product at every outlet in a given city using four vector autoregression models that incorporate the historical retail prices of the product at a target outlet and at competing outlets. The models also include the estimated wholesale price of the product. Three ways of identifying local competitors are considered. The wholesale supplier is that with similar pricing patterns to a target outlet. The proposed models outperform a simple autoregression approach that does not include local competitors or wholesale prices in experiments carried out using data obtained from outlets in five North American cities.

Keywords: price prediction; price forecasting; retail prices; local competitors; vector autoregression; VAR; data mining.

DOI: 10.1504/IJBIDM.2016.076418

International Journal of Business Intelligence and Data Mining, 2016 Vol.11 No.1, pp.19 - 30

Received: 13 Jun 2015
Accepted: 29 Jun 2015
Published online: 06 May 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article