Algorithm 1: Recursive feature elimination algorithm

Input: classification data set D with n features; basic classification algorithm model (such as random forest, logistic regression, etc.); the number of features k ($1\le k\le n$) for iteration stop; the number of features eliminated at each step. 1Initial feature set $F=\{f_1,f_2,\dots ,f_n\}$ 2Training the basic algorithm model:use the F feature set of the data set D to train the basic algorithm model; 3Calculate feature importance: for each feature in F, calculate its impact on model prediction results according to the training results of the model. 4Recursively delete features: delete the least important step features in F; 5Repeat steps 2 to 4 until the required number of features is reached, i.e., $\left|F\right|\le k$; 6Preserve the selected features: use the feature set F to train the final model, and output the feature subset F and model. Output: A feature subset composed of k features.