Compare Linear regression, Decision Tree Regressor, and Random Forest Regressor based on python, a restaurant company on Kaggle as a case

Abstract

The fastest-developed computer language around the world is python because it is easy to use. Most of the linear regression methods currently used for traffic prediction cannot explain the accuracy differences between methods. In this study, the idea of constructing a regression model, decision tree regressor, and random forest regressor is described by analyzing the sales volume, region, and time of a take-out company, and the accuracy difference between the three different models with the same data is compared. It is concluded that the random forest regression volume has the highest accuracy among the three, and the results are obtained by bringing them into the original data for experiments. Although the linear regression model improves the scientific and rational nature of logistics forecasting. It provides a decision basis for enterprises to forecast sales volume. However, if only the linear regression model is used for forecasting, it is still not accurate enough. It needs to be combined with other methods for experiments to further improve accuracy, scientificity, and rationality. In the paper, the basic knowledge of statistics is used to create three different linear regressions through python and find which one is the most accurate. R2 score, MSE, and RMSE are explained. Based on the study, when firms do the prediction of their business or any other projects, if the data set have the linear regression relationship, they could choose the Random Forest through python directly as it is the most precise among them.