ML之xgboost：利用xgboost算法(自带,特征重要性可视化+且作为阈值训练模型)训练mushroom蘑菇数据集(22+1,6513+1611)来预测蘑菇是否毒性(二分类预测)

输出结果

设计思路

核心代码

print('XGB_model.feature_importances_：','\n', XGB_model.feature_importances_)

from matplotlib import pyplot

pyplot.bar(range(len(XGB_model.feature_importances_)), XGB_model.feature_importances_)

from xgboost import plot_importance

plot_importance(XGB_model)

thresholds = sort(XGB_model.feature_importances_)

for thresh in thresholds:

 selection = SelectFromModel(XGB_model, threshold=thresh, prefit=True)

 select_X_train = selection.transform(X_train)

 selection_model = XGBClassifier()

 selection_model.fit(select_X_train, y_train)

 select_X_test = selection.transform(X_test)

 y_pred = selection_model.predict(select_X_test)

 predictions = [round(value) for value in y_pred]

 accuracy = accuracy_score(y_test, predictions)

 print("Thresh=%.3f, n=%d, Accuracy: %.2f%%" % (thresh, select_X_train.shape[1], accuracy*100.0))