无法从 LightGBM 重现 L1 分数

Question

当我运行 LGBM 提前停止时，它会给我对应于其最佳迭代的分数。

当我尝试自己重现这些乐谱时，我得到了不同的数字。

import lightgbm as lgb
from sklearn.datasets import load_breast_cancer
import pandas as pd
from sklearn.metrics import mean_absolute_error
from sklearn.model_selection import KFold


data = load_breast_cancer()
X = pd.DataFrame(data.data)
y = pd.Series(data.target)

lgb_params =  {'boosting_type': 'dart', 'random_state': 42}

folds = KFold(5)

for train_idx, val_idx in folds.split(X):
    X_train, X_valid = X.iloc[train_idx], X.iloc[val_idx]
    y_train, y_valid = y.iloc[train_idx], y.iloc[val_idx]
    model = lgb.LGBMRegressor(**lgb_params, n_estimators=10000, n_jobs=-1)
    model.fit(X_train, y_train,
              eval_set=[(X_valid, y_valid)],
              eval_metric='mae', verbose=-1, early_stopping_rounds=200)
    y_pred_valid = model.predict(X_valid)
    print(mean_absolute_error(y_valid, y_pred_valid))

我期待

valid_0's l1: 0.123608

会匹配我自己根据 mean_absolute_error 的计算结果，但事实并非如此。事实上，这是我输出的顶部：

Training until validation scores don't improve for 200 rounds.
Early stopping, best iteration is:
[631]   valid_0's l2: 0.0515033 valid_0's l1: 0.123608
0.16287265537021847

我使用的是 lightgbm 版本“2.2.1”。

Answer 1

如果您更新 LGBM 版本，您将获得

"UserWarning: Early stopping is not available in dart mode"

详情请参考this issue。您可以做的是使用最佳数量的增强轮次重新训练模型。

results = model.evals_result_['valid_0']['l1']
best_perf = min(results)
num_boost = results.index(best_perf)
print('with boost', num_boost, 'perf', best_perf)    
model = lgb.LGBMRegressor(**lgb_params, n_estimators=num_boost+1, n_jobs=-1)
model.fit(X_train, y_train, verbose=-1)

无法从 LightGBM 重现 L1 分数

Can't reproduce L1-score from LightGBM

python

machine-learning

lightgbm