Research on supermarket replenishment volume prediction based on LSTM and PSO algorithm

Abstract

In order to meet people's health needs and reduce economic losses, the new fresh suppliers carried out real, reliable and targeted market demand analysis so that they formulate automatic pricing and replenishment decision of vegetable commodities with the aim of providing fresh and nutritious vegetable commodities. This paper adopts the apriori model to explore the correlation between vegetables and reflects the relationship between total sales and cost plus pricing by the linear fitting method.This paper further uses the LSTM time series prediction model, time sliding sequence and PSO particle swarm algorithm to predict the various vegetable categories of daily supplement and pricing in the next week, so as to maximize the business super profit. The results showed that the price of each vegetable category had a significant negative correlation with the sales volume and the dynamic prediction of replenishment quantity and pricing data had obvious application effect, which effectively helped the supermarket to make pricing and replenishment decisions.