我正在尝试使用scikit-learn中的f分数作为xgb分类器中的评估指标。这是我的代码:
clf = xgb.XGBClassifier(max_depth=8, learning_rate=0.004,
n_estimators=100,
silent=False, objective='binary:logistic',
nthread=-1, gamma=0,
min_child_weight=1, max_delta_step=0, subsample=0.8,
colsample_bytree=0.6,
base_score=0.5,
seed=0, missing=None)
scores = []
predictions = []
for train, test, ans_train, y_test in zip(trains, tests, ans_trains, ans_tests):
clf.fit(train, ans_train, eval_metric=xgb_f1,
eval_set=[(train, ans_train), (test, y_test)],
early_stopping_rounds=900)
y_pred = clf.predict(test)
predictions.append(y_pred)
scores.append(f1_score(y_test, y_pred))
def xgb_f1(y, t):
t = t.get_label()
return "f1", f1_score(t, y)
但是有一个错误:
Can't handle mix of binary and continuous
最佳答案
问题在于f1_score试图比较非二进制目标和二进制目标，默认情况下，此方法进行二进制平均。来自documentation“平均:字符串，[无，'二进制'(默认)，'micro'，'macro'，'samples'，'weighted']”。
无论如何，错误是说您的预测像[0.001, 0.7889,0.33...]这样是连续的，但是您的目标是二进制[0,1,0...]。因此，如果您知道阈值，建议您在将结果发送到f1_score函数之前对其进行预处理。阈值的通常值为0.5。
评估功能的测试示例。不再输出错误:
def xgb_f1(y, t, threshold=0.5):
t = t.get_label()
y_bin = [1. if y_cont > threshold else 0. for y_cont in y] # binarizing your output
return 'f1',f1_score(t,y_bin)
如@smci所建议的，less_verbose/more_efficiency解决方案可以是:
def xgb_f1(y, t, threshold=0.5):
t = t.get_label()
y_bin = (y > threshold).astype(int) # works for both type(y) == <class 'numpy.ndarray'> and type(y) == <class 'pandas.core.series.Series'>
return 'f1',f1_score(t,y_bin)
关于python - 在XGB中使用F分数，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/35400372/