Research on Fresh Vegetable Market Sales Characteristics and Prediction Model Based on Time Series

Due to the short sales time and storage difficulties of fresh agricultural products, how to price and replenish future fresh products based on historical sales has become an important research topic. This article studies the optimization of sales strategies for fresh vegetables based on time series prediction models and data visualization analysis. Through data collection and preprocessing, detailed sales data from multiple vegetable markets across the country were obtained. Data visualization analysis shows that flowering and leafy vegetables account for about 42% of total vegetable sales, followed by chili plants at 19% and edible fungi at 16%. Taking Hong hu lotus root as an example and using the LSTM time prediction model, it is forecasted that when the supply time is 800 days, the maximum replenishment quantity of crispy lotus roots is about 120,000; when the supply time is 1,000 days, the sales volume will reach the second peak of 100,000. These visualizations provide support for developing targeted sales strategies.