XGBoost自定义测评函数eval_metric

How to custom evaluation metric for XGBoost in Python

# -*- coding: utf-8 -*-
from sklearn.cross_validation import train_test_split
from sklearn.datasets import load_digits
import xgboost as xgb
from sklearn.metrics import matthews_corrcoef
from xgboost import XGBClassifier

THRESHOLD = 0.5

#自定义马修斯相关系数
def evalmcc(preds, dtrain):
    labels = dtrain.get_label()
    return 'MCC', matthews_corrcoef(labels, preds > THRESHOLD)


def evalmcc_min(preds, dtrain):
    labels = dtrain.get_label()
    return 'MCC', -matthews_corrcoef(labels, preds > THRESHOLD)


xgb_params = {
    'seed': 0,
    'colsample_bytree': 0.5,
    'silent': 1,
    'subsample': 0.5,
    'learning_rate': 0.001,
    'objective': 'binary:logistic',
    'max_depth': 2,
    'min_child_weight': 1,
}

if __name__ == "__main__":
    digits = load_digits(n_class=2)
    x_train = digits.data
    y_train = digits.target

    dtrain = xgb.DMatrix(x_train, label=y_train)
    res = xgb.cv(xgb_params, dtrain, num_boost_round=250, nfold=5, seed=0, stratified=True,
                 early_stopping_rounds=25, verbose_eval=5, show_stdv=True, feval=evalmcc, maximize=True)

    clf = XGBClassifier(**xgb_params)

    x_train, x_test, y_train, y_test = train_test_split(x_train, y_train, test_size=0.50, random_state=1337)

    clf.fit(x_train, y_train, eval_set=[(x_test, y_test)], eval_metric=evalmcc_min, early_stopping_rounds=50,
            verbose=True)

http://www.cnblogs.com/silence-gtx/p/5812012.html

https://blog.csdn.net/wl_ss/article/details/78685984

https://www.kaggle.com/c/bosch-production-line-performance/discussion/23909

  • 0
    点赞
  • 7
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值