lightgbm 代码 python

### 使用Python编写的LightGBM模型代码示例

为了创建并训练一个简单的LightGBM模型，可以按照如下方式操作：

首先安装所需的库[^1]：

pip install lightgbm scikit-learn pandas numpy

加载必要的模块并将数据集分为特征和标签两部分:

import lightgbm as lgb
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
import pandas as pd
import numpy as np

# 加载你的数据集到DataFrame中
data = pd.read_csv('your_dataset.csv')

X = data.drop(columns=['target_column'])  # 特征列
y = data['target_column']                  # 目标列

# 将数据划分为训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

定义参数并构建LightGBM的数据结构:

params = {
    'objective': 'binary',          # 对于二分类问题设置目标函数为'binary'
    'metric': ['auc'],              # AUC作为评估指标之一
}

train_data = lgb.Dataset(X_train, label=y_train)
test_data = lgb.Dataset(X_test, label=y_test, reference=train_data)

训练模型并保存最佳迭代次数的结果:

bst = lgb.train(params,
                train_data,
                valid_sets=[test_data],
                num_boost_round=999,
                early_stopping_rounds=10)

best_iteration = bst.best_iteration
print(f"Best iteration: {best_iteration}")

利用训练好的模型预测新样本，并计算准确率得分:

predictions_probabilities = bst.predict(X_test, num_iteration=best_iteration)
predicted_classes = (predictions_probabilities >= 0.5).astype(int)

accuracy = accuracy_score(y_test, predicted_classes)
print(f"Accuracy score on testing set is {accuracy:.4f}")