Production model based on the combined prediction of time series and XGBoost neural networks

This paper discusses the production problem of multi-variety materials. With the continuous development of economic globalization, the production and operation mode of many varieties of small batch materials has been loved by the company, and the market demand for materials determines the quantity of our production materials. In the code of small-batch materials in the given sample, we first use Python data to analyze the trend level of the material, establish the entropy right Topsis model to objectively score the material, and screen out the six materials that should be paid attention to in the materials that meet the requirements. Secondly, with the week as the time unit, the time series and XGBoost neural network are used to establish a weekly prediction model to predict the six material demand, and the reciprocal mean square error method is used to combine the two models to ensure the accuracy of the model and improve the credibility of the model, to build the combined prediction model.