我的第一个多类分类.我有值X和Y.Y有5个值[0,1,2,3,4].但我得到这个"不支持多类格式".理解我需要xgb_params中的num_class,但是如果我想要'num_class':range(0,5,1)而不是获取estimator XGBClassifier的无效参数num_class.
xgb_model = xgb.XGBClassifier(objective='multi:softmax')
xgb_params = [
{
"n_estimators": range(50, 501, 50),
}
]
cv = cross_validation.StratifiedShuffleSplit(y_train, n_iter=5,
test_size=0.3, random_state=42)
xgb_grid = grid_search.GridSearchCV(xgb_model, xgb_params, scoring='roc_auc', cv=cv, n_jobs=-1, verbose=3)
xgb_grid.fit(X_train, y_train)
这个值的例子:
X Y
-1.35173485 1.50224188 2.04951167 0.43759658 0.24381777 2
2.81047260 1.31259056 1.39265240 0.16384002 0.65438366 3
2.32878809 -1.92845940 -2.06453246 0.73132270 0.11771229 2
-0.12810555 -2.07268765 -2.40760215 0.97855042 0.11144164 1
1.88682063 0.75792329 -0.09754671 0.46571931 0.62111648 2
-1.09361266 1.74758304 2.49960891 0.36679883 0.88895562 2
0.71760095 -1.30711698 -2.15681966 0.33700593 0.07171119 2
4.60060308 -1.60544855 -1.88996123 0.94500124 …