Fig. 4
From: Measuring China’s Policy Stringency on Climate Change for 1954–2022
The Model Performance. This diagram compares the residual distribution of various models: Random Forest (RF), eXtreme gradient boosting with tree booster (GBM), cubist regression model (Cubist), a bagging wrapper for multivariate adaptive regression splines via the earth function (BagEarth), Support Vector Machine (SVM), Lasso Regression (lasso), Elastic Net (Elastic), and Linear Model (LM). The x-axis shows residuals, and the y-axis displays the percentage of data with residuals at least as large. Steeper curves indicate fewer large residuals, signifying better model performance.
