Pearson Correlation Analysis of Vegetable Categories based on K-L-ISODATA Clustering Algorithm

In order to make the superstore's revenue as much as possible and develop stably, firstly, we carry out descriptive analysis and Pearson correlation analysis on the daily sales volume of each vegetable category, and then we use K-L-ISODATA clustering algorithm for clustering of the daily sales volume change trend, and we get 6 categories of vegetable categories with the same change trend. Next, an optimisation model was established with the objective function of superstore revenue. The cost of each vegetable category was predicted by LSTM time series on July 1-7, 2023, and the total daily replenishment volume and pricing strategy of each vegetable category in the coming week were obtained by simulated annealing algorithm with an optimisation solution. Then use entropy weight method TOPSIS to carry out comprehensive evaluation of single products, selected the comprehensive score ranked the top 33 single products, to take the cycle will be the first 27-33 in turn for the superstore revenue value, the largest revenue combination for the top 30 of each single product, and according to the same steps in the previous optimisation model building.