Sales prediction method and system based on prophet model and big data

Abstract

The invention relates to the technical field of big data mining, and provides an industry big data sales volume prediction method based on a Prophet algorithm for solving the problem of low prediction accuracy of an existing sales volume prediction method in the automobile industry, and the method comprises the steps: carrying out the analysis of holidays and festivals, events and the like on historical sales volume data based on a facebook prophet algorithm, determining influence parameters, adopting historical sales volume data to carry out training test on the proret model to obtain an optimal model, and predicting the sales volume to obtain a first predicted value; performing moving average, deformation, dimension reduction, lag and the like on the influence factors, performing machine learning model (neural network, xgboost and the like) training test by adopting historical sales volume data and all influence factor data to obtain an optimal machine learning model, and predicting the sales volume by utilizing the optimal machine learning model and a factor prediction value to obtain a second prediction value; and obtaining a final sales volume predicted value according to the sales volume predicted value I and the sales volume predicted value II. The method is suitable for predicting the sales volume of the automobile products.

Description

technical field [0001] The invention relates to the field of big data technology, in particular to a method for predicting industry product sales based on data mining technology. Background technique [0002] In order to better control inventory, reduce capital pressure and fundamentally control costs, improve market adaptability, and increase corporate profits, effective sales forecasting methods are urgently needed, especially in the automotive industry, where capital occupation costs are high and remodeling cycles are long. Inventory pressure is high. At present, the automobile industry is in the transition stage from the early stage of popularization to the later stage of popularization. After the end of unilateral growth, and a large number of uncompetitive products are marginalized by the market, sales forecast can promote the matching of production and sales on the one hand, and identify the Changes in market trends provide support for product changes. Sales forecas...

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is: in view of the existing product sales forecasting system, factors such as the influence of events in the sales of the automobile industry, the long remodeling period, and the large amount of capital occupied are not considered, and the prediction accuracy rate is low by using a general method. problem, propose a sales forecast method for the automobile industry based on big data mining technology

Images