Wavelet Neural Network Based Study of Merchandise Replenishment and Pricing

Abstract

Merchandise replenishment pricing strategy is the most critical step in a merchant's sales decision. The formulation of reasonable and accurate replenishment strategy is of great significance for controlling the cost of sales and improving economic efficiency. In order to reasonably formulate the replenishment strategy and price, this paper is based on the wavelet neural network theory, increased the forecast model for the nonlinear relationship between the commodity sales portfolio, combined with the Spearman correlation coefficient to fully explore the correlation between the commodity sales portfolio and its impact on the number of commodities for sale, the use of cost plus pricing method and the rate of loss of commodities with the loss of value of the commodities over time, to deal with, construct a dynamic pricing and replenishment model of goods. The model is more capable of handling the non-relationships in merchandising and is more generalizable than the traditional time series forecasting model. Compared with previous studies, this model better handles the relationship between merchandise sales combinations through wavelet transform, solves the limitation that time series forecasting cannot fully consider the correlation relationship between merchandise sales combinations, and provides a new solution for the treatment of sales combinations in merchandise sales forecasting.