A Multiple Regression Model National Grain Consumption Prediction Method

Abstract

A multivariate regression model national grain consumption prediction method based on interpolation group sliding center of gravity, the method includes the following steps in turn: (1) Obtaining original data; (2) Obtaining important influencing factors; (3) Grain consumption and two The important influencing factors are preprocessed to obtain the preprocessed data; (4) Find the sliding center of gravity of the preprocessed data; (5) Input the data obtained in step (4) into the multiple regression forecasting model to obtain the preliminary forecast result of grain consumption; ( 6) Carry out the inverse calculation of the center of gravity of the sliding data on the prediction result obtained in step (5) to obtain the actual predicted value of the required grain consumption y' 1 (t) . This method can smooth the prediction error, and the added interpolation preprocessing can compensate for the decrease of the center of gravity data in the process of grouping sliding center of gravity calculation, and at the same time increase the training data of the prediction model, which significantly improves the prediction accuracy of the multiple regression model.

Description

technical field [0001] The invention belongs to the technical field of ration consumption forecasting, and in particular relates to a method for forecasting national grain consumption based on a multivariate regression model of interpolation grouping sliding center of gravity. Background technique [0002] Grain consumption is an important part of the national grain consumption. Data analysis shows that with the acceleration of urbanization level construction, the grain consumption presents a decreasing trend year by year, and there are obvious differences in urban and rural ration consumption trends. At present, many scholars have proposed methods such as EMM model method, time series extrapolation method, panel data estimation, statistical analysis and econometric analysis to predict grain consumption, and review the existing methods. There are historical data anomalies in grain consumption prediction methods It is very destructive to the prediction results, the lack of or...

AI Technical Summary

Problems solved by technology

At present, many scholars have proposed methods such as EMM model method, time series extrapolation method, panel data estimation, statistical analysis and econometric analysis to predict grain consumption, and review the existing methods. There are historical data anomalies in grain consumption prediction methods It is very destructive to the prediction results, the lack of original data leads to large training errors of the prediction model, and the prediction of a single dependent variable.

Images