探索AI人工智能领域数据挖掘的无限可能

    import pandas as pd
    import numpy as np
    from sklearn.preprocessing import StandardScaler, OneHotEncoder
    from sklearn.impute import SimpleImputer
    from sklearn.compose import ColumnTransformer
    from sklearn.pipeline import Pipeline
    
    # 示例数据集
    data = {
    'age': [25, 30, np.nan, 35, 40],
    'income': [50000, np.nan, 70000, 80000, 90000],
    'gender': ['M', 'F', 'M', 'F', 'M'],
    'purchased': ['yes', 'no', 'no', 'yes', 'yes']
    }
    
    df = pd.DataFrame(data)
    
    # 定义预处理步骤
    numeric_features = ['age', 'income']
    numeric_transformer = Pipeline(steps=[
    ('imputer', SimpleImputer(strategy='median')),
    ('scaler', StandardScaler())])
    
    categorical_features = ['gender']
    categorical_transformer = Pipeline(steps=[
    ('imputer', SimpleImputer(strategy='constant', fill_value='missing')),
    ('onehot', OneHotEncoder(handle_unknown='ignore'))])
    
    preprocessor = ColumnTransformer(
    transformers=[
        ('num', numeric_transformer, numeric_features),
        ('cat', categorical_transformer, categorical_features)])
    
    # 应用预处理
    X = df.drop('purchased', axis=1)
    y = df['purchased']
    
    X_preprocessed = preprocessor.fit_transform(X)
    print("预处理后的数据:\n", X_preprocessed)
    
    
    python
    
    
![](https://ad.itadn.com/c/weblog/blog-img/images/2025-08-16/uh1Rq4ArUZbQleIJFjD0v5SiKdM7.png)

    from sklearn.tree import DecisionTreeClassifier
    from sklearn.model_selection import train_test_split
    from sklearn.metrics import accuracy_score
    
    # 使用预处理后的数据
    X_train, X_test, y_train, y_test = train_test_split(X_preprocessed, y, test_size=0.2, random_state=42)
    
    # 创建决策树模型
    tree_clf = DecisionTreeClassifier(max_depth=3, random_state=42)
    tree_clf.fit(X_train, y_train)
    
    # 预测和评估
    y_pred = tree_clf.predict(X_test)
    print("决策树准确率:", accuracy_score(y_test, y_pred))
    
    
    python
    
    
![](https://ad.itadn.com/c/weblog/blog-img/images/2025-08-16/WrPRafDV1OuQKoF3BUzcZAtiv62s.png)

    from sklearn.ensemble import RandomForestClassifier
    
    rf_clf = RandomForestClassifier(n_estimators=100, max_depth=3, random_state=42)
    rf_clf.fit(X_train, y_train)
    
    y_pred_rf = rf_clf.predict(X_test)
    print("随机森林准确率:", accuracy_score(y_test, y_pred_rf))
    
    
    python
    
    

    from sklearn.cluster import KMeans
    import matplotlib.pyplot as plt
    
    # 生成示例数据
    from sklearn.datasets import make_blobs
    X, y = make_blobs(n_samples=300, centers=4, cluster_std=0.60, random_state=0)
    
    # 应用K-Means
    kmeans = KMeans(n_clusters=4)
    kmeans.fit(X)
    y_kmeans = kmeans.predict(X)
    
    # 可视化结果
    plt.scatter(X[:, 0], X[:, 1], c=y_kmeans, s=50, cmap='viridis')
    centers = kmeans.cluster_centers_
    plt.scatter(centers[:, 0], centers[:, 1], c='red', s=200, alpha=0.5)
    plt.title("K-Means聚类结果")
    plt.show()
    
    
    python
    
    
![](https://ad.itadn.com/c/weblog/blog-img/images/2025-08-16/8FpGZQBdNPCq2bu7WwzRDaHIoKkL.png)

    conda create -n data_mining python=3.8
    conda activate data_mining
    conda install numpy pandas scikit-learn matplotlib seaborn jupyter
    
    
    bash

    # 导入必要库
    import pandas as pd
    import numpy as np
    import matplotlib.pyplot as plt
    import seaborn as sns
    from sklearn.model_selection import train_test_split, GridSearchCV
    from sklearn.preprocessing import StandardScaler
    from sklearn.ensemble import RandomForestClassifier
    from sklearn.metrics import classification_report, confusion_matrix, roc_auc_score
    from imblearn.over_sampling import SMOTE
    
    # 加载数据
    data = pd.read_csv('customer_churn.csv')
    
    # 探索性数据分析
    print(data.head())
    print(data.info())
    print(data.describe())
    
    # 可视化特征分布
    plt.figure(figsize=(12,6))
    data['Churn'].value_counts().plot(kind='bar')
    plt.title('Churn Distribution')
    plt.show()
    
    # 数据预处理
    # 处理缺失值
    data.fillna(data.median(), inplace=True)
    
    # 特征选择
    X = data.drop(['customerID', 'Churn'], axis=1)
    y = data['Churn']
    
    # 处理分类变量
    X = pd.get_dummies(X, drop_first=True)
    
    # 处理类别不平衡
    smote = SMOTE(random_state=42)
    X_res, y_res = smote.fit_resample(X, y)
    
    # 数据分割
    X_train, X_test, y_train, y_test = train_test_split(X_res, y_res, test_size=0.2, random_state=42)
    
    # 特征缩放
    scaler = StandardScaler()
    X_train = scaler.fit_transform(X_train)
    X_test = scaler.transform(X_test)
    
    # 模型训练
    rf = RandomForestClassifier(random_state=42)
    param_grid = {
    'n_estimators': [100, 200],
    'max_depth': [None, 10, 20],
    'min_samples_split': [2, 5]
    }
    
    grid_search = GridSearchCV(rf, param_grid, cv=5, scoring='roc_auc', n_jobs=-1)
    grid_search.fit(X_train, y_train)
    
    # 最佳模型
    best_rf = grid_search.best_estimator_
    
    # 模型评估
    y_pred = best_rf.predict(X_test)
    y_proba = best_rf.predict_proba(X_test)[:,1]
    
    print("分类报告:")
    print(classification_report(y_test, y_pred))
    
    print("混淆矩阵:")
    print(confusion_matrix(y_test, y_pred))
    
    print("ROC AUC得分:", roc_auc_score(y_test, y_proba))
    
    # 特征重要性
    feature_importance = pd.DataFrame({
    'feature': X.columns,
    'importance': best_rf.feature_importances_
    }).sort_values('importance', ascending=False)
    
    plt.figure(figsize=(12,8))
    sns.barplot(x='importance', y='feature', data=feature_importance.head(10))
    plt.title('Top 10 Important Features')
    plt.show()
    
    
    python
    
    
![](https://ad.itadn.com/c/weblog/blog-img/images/2025-08-16/T9wqXGaVgyUQ8m2ipN6sv0Suz1Bl.png)

 * 处理缺失值：用中位数填充数值特征的缺失值
 * 特征选择：删除无关特征(ID列和目标列)
 * 分类变量处理：使用独热编码转换分类变量
 * 类别不平衡处理：使用SMOTE算法生成合成样本平衡类别

 * 使用随机森林作为基础模型
 * 通过网格搜索(GridSearchCV)寻找最佳超参数组合
 * 采用5折交叉验证评估模型性能
 * 使用ROC AUC作为主要评估指标

 * 输出分类报告(精确率、召回率、F1分数)
 * 显示混淆矩阵
 * 计算ROC AUC得分
 * 可视化特征重要性

 * 类别不平衡是常见问题，SMOTE是有效的解决方法之一
 * 随机森林能自动处理特征间的非线性关系
 * 特征重要性分析有助于业务理解和模型解释

 * 客户细分和个性化推荐
 * 购物篮分析(关联规则挖掘)
 * 需求预测和库存优化

 * 信用评分和风险评估
 * 欺诈检测
 * 算法交易

 * 疾病预测和诊断辅助
 * 药物发现
 * 医疗资源优化

 * 预测性维护
 * 质量控制
 * 供应链优化

 * 客户流失预测(如我们案例所示)
 * 网络优化
 * 异常检测

 * 情感分析
 * 影响力用户识别
 * 内容推荐

 * 交通流量预测
 * 公共安全分析
 * 资源分配优化