预测用户收入是否大于50k

我们建立xbgoost模型得到如下性能指标

xgboost自带特征重要性

Skater Feature Importance

Skater Partial Dependency

Skater Partial Dependency

Skater 2 Ways Partial Dependency

Skater local interpretation

Skater local interpretation

Skater surrogate tree

SHAP Local Interpretation

SHAP Local Interpretation

SHAP Local Batch nterpretation

SHAP Global Interpretation

SHAP Global Interpretation

SHAP Partial Dependency

SHAP Partial Dependency

SHAP 2 ways Partial Dependency

SHAP 2 ways Partial Dependency

SHAP 2 ways Partial Dependency