用python写代码

为了帮助您编写 Python 代码以完成实验三的任务，我将提供一个详细的步骤指南，包括数据预处理、模型构建、评估和结果分析。我们将使用 `scikit-learn` 库来实现这些任务。

### 实验步骤

#### 1. 导入必要的库
首先，导入所有必要的库：

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import accuracy_score, classification_report
from sklearn.ensemble import VotingClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.svm import SVC
from sklearn.datasets import load_iris

#### 2. 加载和预处理数据

##### Glass 数据集
假设 `glass.data` 和 `glass.names` 文件已经下载并保存在当前目录下。

# 加载玻璃数据集
column_names = ['RI', 'Na', 'Mg', 'Al', 'Si', 'K', 'Ca', 'Ba', 'Fe', 'Type']
data = pd.read_csv('glass.data', names=column_names)

# 分离特征和标签
X = data.drop('Type', axis=1)
y = data['Type']

# 将多类别问题转换为二分类问题
y_binary = (y == 1).astype(int)  # 假设类别1为正类

# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y_binary, test_size=0.3, random_state=42)

# 标准化特征
scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)

##### Leukemia 数据集
假设 `leukemia.data` 文件已经下载并保存在当前目录下。

# 加载白血病数据集
data_leukemia = pd.read_csv('leukemia.data', sep='\t')
X_leukemia = data_leukemia.iloc[:, 1:].values
y_leukemia = data_leukemia.iloc[:, 0].values

# 划分训练集和测试集
X_train_leukemia, X_test_leukemia, y_train_leukemia, y_test_leukemia = train_test_split(
    X_leukemia, y_leukemia, test_size=0.3, random_state=42)

# 标准化特征
scaler_leukemia = StandardScaler()
X_train_leukemia_scaled = scaler_leukemia.fit_transform(X_train_leukemia)
X_test_leukemia_scaled = scaler_leukemia.transform(X_test_leukemia)

#### 3. 构建和评估基分类器

##### Glass 数据集

# 定义基分类器
clf1 = LogisticRegression(max_iter=1000)
clf2 = DecisionTreeClassifier()
clf3 = SVC(probability=True)

# 训练基分类器
for clf in [clf1, clf2, clf3]:
    clf.fit(X_train_scaled, y_train)
    y_pred = clf.predict(X_test_scaled)
    print(f"Classifier: {clf.__class__.__name__}")
    print("Accuracy:", accuracy_score(y_test, y_pred))
    print(classification_report(y_test, y_pred))

##### Leukemia 数据集

# 训练基分类器
for clf in [clf1, clf2, clf3]:
    clf.fit(X_train_leukemia_scaled, y_train_leukemia)
    y_pred = clf.predict(X_test_leukemia_scaled)
    print(f"Classifier: {clf.__class__.__name__}")
    print("Accuracy:", accuracy_score(y_test_leukemia, y_pred))
    print(classification_report(y_test_leukemia, y_pred))

#### 4. 构建和评估集成分类器

##### Soft 和 Hard 投票集成

# 定义投票集成分类器
voting_clf_hard = VotingClassifier(estimators=[('lr', clf1), ('dt', clf2), ('svc', clf3)], voting='hard')
voting_clf_soft = VotingClassifier(estimators=[('lr', clf1), ('dt', clf2), ('svc', clf3)], voting='soft')

# 训练投票集成分类器
voting_clf_hard.fit(X_train_scaled, y_train)
voting_clf_soft.fit(X_train_scaled, y_train)

# 评估投票集成分类器
for clf, label in zip([voting_clf_hard, voting_clf_soft], ['Hard Voting', 'Soft Voting']):
    y_pred = clf.predict(X_test_scaled)
    print(f"Ensemble Method: {label}")
    print("Accuracy:", accuracy_score(y_test, y_pred))
    print(classification_report(y_test, y_pred))

##### 权重加权软投票集成

# 计算每个基分类器的交叉验证得分
weights = []
for clf in [clf1, clf2, clf3]:
    scores = cross_val_score(clf, X_train_scaled, y_train, cv=5)
    weights.append(np.mean(scores))

# 定义加权软投票集成分类器
weighted_voting_clf = VotingClassifier(estimators=[('lr', clf1), ('dt', clf2), ('svc', clf3)], voting='soft', weights=weights)

# 训练加权软投票集成分类器
weighted_voting_clf.fit(X_train_scaled, y_train)

# 评估加权软投票集成分类器
y_pred = weighted_voting_clf.predict(X_test_scaled)
print("Ensemble Method: Weighted Soft Voting")
print("Accuracy:", accuracy_score(y_test, y_pred))
print(classification_report(y_test, y_pred))

#### 5. 结果分析
您可以根据上述输出结果进行分析，比较不同基分类器和集成方法的性能，特别是关注准确率、召回率和F1分数等指标。

希望这些代码和步骤能帮助您顺利完成实验三的任务！如果有任何问题或需要进一步的帮助，请随时告诉我。