Product Business Decision Making Based on Simulated Annealing Algorithm and Time Series Analysis

The aim of this paper is to optimize the dynamic pricing and replenishment strategy of commodities by establishing a decision model based on simulated annealing algorithm and time series analysis. According to existing surveys, some retailers try to manage inventory through simple Cost-Plus Pricing or static replenishment strategies, but often ignore the complex correlation between products and the volatility of market demand, making it difficult to adapt to rapid changes in the market environment. Traditional retail pricing and inventory management approaches fail to take full advantage of intelligent optimization algorithms, resulting in excess or shortage of inventory, unreasonable pricing, and unoptimized resource allocation. The study first analyzes the descriptive statistics of the sales data of various types of commodities to reveal the volatility of sales and its correlation among different categories. Subsequently, Pearson correlation analysis was applied to confirm the existence of significant positive correlations between various categories of merchandise. Based on the data of recent years, a time series model is established by R/S rescaled polar method to predict the future daily cost, and the optimal replenishment quantity and pricing strategy are derived by combining with the calculation of wastage rate. Finally, this paper constructs an optimization model by Monte Carlo method to screen out a variety of sellable single products. The optimized model can help retailers to stand out from the competition in the market and improve the allocation efficiency of resources in general.