Research on Default Prediction for Credit Card Users Based on XGBoost-LSTM Model

Table 4

Machine learning model data processing methods and model settings.

Algorithm	Data processing	Model settings
K neighbors	Normalization, missing value supplement	n_neighbors = 10
SVM	Normalization, missing value supplement	kernel = “rbf,” max_iter = 500
DecisionTree	—	max_depth = 4
RandomForest	—	max_depth = 4
AdaBoost	—	max_depth = 4
XGBoost	—	max_depth = 4