Retail Commodity Sale Forecast Model Based on Data Mining

Abstract

In terms of the retail commodity sale forecast, people did more in particular aspect with commodity's single sale attribute such as the sale volume, the sale money, the season factor, but all has not considered the most important factor-profit, the profit is the key factor of retail enterprises winning the survival and development. However, such a one-sided analysis is not conducive to assist the managers understand the overall situation of retail sales, and make the decision the sale and the inventory. So this paper firstly selected the profit ratio which was on behalf of commodity profit element and several other key sale attributes including the season ratio and the sale volume to establish the SPV Model, secondly done commodity sale state segmentation based on the SPV Model with ID3 decision tree algorithm, And on this basis we predicted the sale state of the commodity at some future time, Finally, we compared and analysis the results of the SPV Model, the Season Model and the Markov Model through experiments, and get the conclusion that the SPV Model can reach higher correctness than the other two.