tree_shap Namespace Reference

Variables
	data = fetch_california_housing()
	X = data.data
	y = data.target
int	num_round = 500
dict	param
	dtrain = xgb.DMatrix(X, label=y, feature_names=data.feature_names)
	model = xgb.train(param, dtrain, num_round)
	shap_values = model.predict(dtrain, pred_contribs=True)
	shap_interaction_values = model.predict(dtrain, pred_interactions=True)
	explainer = shap.TreeExplainer(model)
	feature_names
	matplotlib
	plot_type

Detailed Description

Use GPU to speedup SHAP value computation
=========================================

Demonstrates using GPU acceleration to compute SHAP values for feature importance.

Variable Documentation

param

dict tree_shap.param

Initial value:

=  {
    "eta": 0.05,
    "max_depth": 10,
    "tree_method": "hist",
    "device": "cuda",
}