### 数据集分类方法及Python代码实现 我们将使用三种分类方法对给定的数据集进行分类，其中包括一种集成方法。以下是详细的Python代码实现： #### 方法一：逻辑回归 (Logistic Regression) ```python import pandas as pd from sklearn.model_selection import train_test_split, cross_val_score from sklearn.preprocessing import StandardScaler, LabelEncoder from sklearn.linear_model import LogisticRegression from sklearn.metrics import accuracy_score # 加载数据 df = pd.read_csv('dataming.csv') # 数据预处理 le = LabelEncoder() df['VisitorType'] = le.fit_transform(df['VisitorType']) df['Weekend'] = df['Weekend'].astype(int) X = df.drop(columns=['Weekend']) y = df['Weekend'] # 特征选择（这里简化为全部特征） # 实际应用中应选择相关性较高的特征 X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42) # 数据标准化 scaler = StandardScaler() X_train_scaled = scaler.fit_transform(X_train) X_test_scaled = scaler.transform(X_test) # 模型训练 logreg = LogisticRegression(max_iter=1000) logreg.fit(X_train_scaled, y_train) # 预测 y_pred = logreg.predict(X_test_scaled) # 评估模型 accuracy = accuracy_score(y_test, y_pred) print(f'Logistic Regression Accuracy: {accuracy:.4f}') cross_val_scores = cross_val_score(logreg, X_train_scaled, y_train, cv=5) print(f'Cross Validation Scores: {cross_val_scores.mean():.4f} (+/- {cross_val_scores.std() * 2:.4f})') ``` #### 方法二：支持向量机 (Support Vector Machine) ```python from sklearn.svm import SVC # 使用支持向量机 svc = SVC(kernel='rbf', probability=True) svc.fit(X_train_scaled, y_train) # 预测 y_pred_svc = svc.predict(X_test_scaled) # 评估模型 accuracy_svc = accuracy_score(y_test, y_pred_svc) print(f'SVM Accuracy: {accuracy_svc:.4f}') cross_val_scores_svc = cross_val_score(svc, X_train_scaled, y_train, cv=5) print(f'SVM Cross Validation Scores: {cross_val_scores_svc.mean():.4f} (+/- {cross_val_scores_svc.std() * 2:.4f})') ``` #### 方法三：随机森林 (Random Forest, 集成方法) ```python from sklearn.ensemble import RandomForestClassifier # 使用随机森林 rf = RandomForestClassifier(n_estimators=100, random_state=42) rf.fit(X_train, y_train) # 预测 y_pred_rf = rf.predict(X_test) # 评估模型 accuracy_rf = accuracy_score(y_test, y_pred_rf) print(f'Random Forest Accuracy: {accuracy_rf:.4f}') cross_val_scores_rf = cross_val_score(rf, X_train, y_train, cv=5) print(f'Random Forest Cross Validation Scores: {cross_val_scores_rf.mean():.4f} (+/- {cross_val_scores_rf.std() * 2:.4f})') ``` ### 完整代码整合 为了方便执行，将所有代码整合在一起，并添加必要的导入和注释： ```python import pandas as pd from sklearn.model_selection import train_test_split, cross_val_score from sklearn.preprocessing import StandardScaler, LabelEncoder from sklearn.linear_model import LogisticRegression from sklearn.svm import SVC from sklearn.ensemble import RandomForestClassifier from sklearn.metrics import accuracy_score # 加载数据 df = pd.read_csv('dataming.csv') # 数据预处理 le = LabelEncoder() df['VisitorType'] = le.fit_transform(df['VisitorType']) df['Weekend'] = df['Weekend'].astype(int) X = df.drop(columns=['Weekend']) y = df['Weekend'] # 分割数据集 X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42) # 数据标准化 scaler = StandardScaler() X_train_scaled = scaler.fit_transform(X_train) X_test_scaled = scaler.transform(X_test) # 方法一：逻辑回归 logreg = LogisticRegression(max_iter=1000) logreg.fit(X_train_scaled, y_train) y_pred_logreg = logreg.predict(X_test_scaled) accuracy_logreg = accuracy_score(y_test, y_pred_logreg) print(f'Logistic Regression Accuracy: {accuracy_logreg:.4f}') cross_val_scores_logreg = cross_val_score(logreg, X_train_scaled, y_train, cv=5) print(f'Logistic Regression Cross Validation Scores: {cross_val_scores_logreg.mean():.4f} (+/- {cross_val_scores_logreg.std() * 2:.4f})') # 方法二：支持向量机 svc = SVC(kernel='rbf', probability=True) svc.fit(X_train_scaled, y_train) y_pred_svc = svc.predict(X_test_scaled) accuracy_svc = accuracy_score(y_test, y_pred_svc) print(f'SVM Accuracy: {accuracy_svc:.4f}') cross_val_scores_svc = cross_val_score(svc, X_train_scaled, y_train, cv=5) print(f'SVM Cross Validation Scores: {cross_val_scores_svc.mean():.4f} (+/- {cross_val_scores_svc.std() * 2:.4f})') # 方法三：随机森林（集成方法） rf = RandomForestClassifier(n_estimators=100, random_state=42) rf.fit(X_train, y_train) y_pred_rf = rf.predict(X_test) accuracy_rf = accuracy_score(y_test, y_pred_rf) print(f'Random Forest Accuracy: {accuracy_rf:.4f}') cross_val_scores_rf = cross_val_score(rf, X_train, y_train, cv=5) print(f'Random Forest Cross Validation Scores: {cross_val_scores_rf.mean():.4f} (+/- {cross_val_scores_rf.std() * 2:.4f})') ``` ### 知识点 1. **逻辑回归**：一种广义线性模型，用于解决二分类问题，通过 logistic 函数将线性组合映射到 [0, 1] 区间。 - 公式：$p = \frac{1}{1 + e^{-z}}$，其中 $z = w^T x + b$ 2. **支持向量机**：通过最大化间隔来找到最优超平面，适用于高维空间分类。 - 核函数（如 RBF 核）用于将输入空间映射到高维特征空间。 3. **随机森林**：由多棵决策树组成的集成学习方法，通过投票机制输出最终结果。 - 随机森林通过 bootstrap 抽样和随机特征选择构建多棵树，从而减少过拟合并提高准确性。 ### 注意事项 - 在实际应用中，建议根据数据特点调整参数，并进行更详细的特征工程。 - 代码中假设所有列都是有效的特征，实际应用中可能需要进一步选择或处理特征。 - `dataming.csv` 文件路径需要根据实际情况调整。