更新 function.py

function.py

@@ -1,115 +1,134 @@
+# function.py
 import numpy as np
+import pandas as pd
+from sklearn.preprocessing import StandardScaler
+from sklearn.model_selection import train_test_split, KFold
 
-# 这里填写学号和姓名
-id = "202210218225"
-name = "周水源"
+"""
+实验说明:
+- 自行编写实现数据加载与预测病马死亡的对率回归算法。
 
+注意事项:
+- 在 function.py 中实现对率回归算法。
+- 可以自行导入其他库，需要在 requirements.txt 添加依赖，若未在 requirements.txt 中添加使用到的依赖，则无法自动评测，成绩为零。
+- 自动评测脚本仅会加载 train_logreg、predict_logreg 和 load_data 函数，请勿修改函数名。
+
+测评环境:
+- CPU: Intel Xeon(R) CPU E3-1231 v3 @ 3.40GHz
+- OS: Ubuntu 22.04.4 LTS
+- Python: 3.10.12
+- numpy: 2.1.1
+"""
+
+
+"""
+请在这里填写学号和姓名，不填写即成绩无效。
+"""
+id = "202210423209"
+name = "万子玉"
+
+# 这里写自己的代码实现
 def sigmoid(z):
-    """
-    Sigmoid 函数，将输入值映射到 (0, 1) 的范围。
-    """
-    return 1 / (1 + np.exp(-z))
+    return 1.0 / (1 + np.exp(-z))
 
-def train_logreg(X, y, learning_rate=0.01, iterations=1000, lambda_reg=0.1):
-    """
-    对率回归训练函数，加入了 L2 正则化。
+def compute_gradient(X_batch, y_batch, weights, lambda_reg):
+    z = np.dot(X_batch, weights)
+    h = sigmoid(z)
+    gradient = np.dot(X_batch.T, (h - y_batch)) / len(y_batch)
+    gradient += (lambda_reg / len(y_batch)) * weights  # L2正则化
+    return gradient
 
-    Args:
-    - X (numpy array): 特征数据。
-    - y (numpy array): 标签数据。
-    - learning_rate (float): 学习率，默认值为 0.01。
-    - iterations (int): 梯度下降迭代次数，默认值为 1000。
-    - lambda_reg (float): 正则化强度，默认值为 0.1。
+def train_logreg(X, y, learning_rate=0.001, iterations=10000, batch_size=64, lambda_reg=0.1):
+    m, n = X.shape
+    weights = np.random.randn(n) * np.sqrt(2. / n)  # He初始化
+    best_weights = weights.copy()
+    best_accuracy = 0
 
-    Returns:
-    - numpy array: 训练得到的权重。
-    """
-    # 初始化权重
-    weights = np.zeros(X.shape[1])
+    for _ in range(iterations):
+        indices = np.random.permutation(m)
+        X_shuffled = X[indices]
+        y_shuffled = y[indices]
 
-    # 梯度下降
-    for i in range(iterations):
-        # 计算预测值
-        predictions = sigmoid(np.dot(X, weights))
+        for j in range(0, m, batch_size):
+            X_batch = X_shuffled[j:j + batch_size]
+            y_batch = y_shuffled[j:j + batch_size]
+            gradient = compute_gradient(X_batch, y_batch, weights, lambda_reg)
+            weights -= learning_rate * gradient
 
-        # 计算误差
-        error = predictions - y
+        # 计算训练集准确率
+        train_accuracy = accuracy(X, y, weights)
+        if train_accuracy > best_accuracy:
+            best_accuracy = train_accuracy
+            best_weights = weights.copy()
 
-        # 计算梯度，并加入正则化项
-        gradient = np.dot(X.T, error) / len(y) + lambda_reg * weights / len(y)
-        weights -= learning_rate * gradient
+        # 动态调整学习率
+        if _ % 100 == 0 and train_accuracy > 0.95:  # 达到高准确率时减小学习率
+            learning_rate *= 0.9
 
-        # 每 100 次迭代输出一次损失
-        if i % 100 == 0:
-            loss = -np.mean(y * np.log(predictions) + (1 - y) * np.log(1 - predictions)) + (lambda_reg / (2 * len(y))) * np.sum(weights**2)
-            print(f"Iteration {i}, Loss: {loss:.4f}")
+    return best_weights
 
-    return weights
-
-def predict_logreg(inX, weights):
-    """
-    对率回归预测函数。
-
-    Args:
-    - inX (numpy array): 输入待分类的样本。
-    - weights (numpy array): 训练得到的权重。
-
-    Returns:
-    - int: 预测的分类标签 (0 或 1)。
-    """
-    # 计算预测概率
-    prob = sigmoid(np.dot(inX, weights))
-    # 将概率转为标签
-    return 1 if prob >= 0.5 else 0
+def predict_logreg(X, weights):
+    z = np.dot(X, weights)
+    h = sigmoid(z)
+    return np.where(h >= 0.5, 1, 0)
 
 def load_data(filepath):
-    """
-    从文件中加载数据，并处理 '?' 为每列的均值。
-    """
-    data = []
-    labels = []
-    
-    with open(filepath, 'r') as file:
-        for line in file:
+    data, labels = [], []
+    with open(filepath, 'r') as f:
+        for line in f:
             line = line.strip().split()
-            if not line:
-                continue
-            # 将特征转换为浮点型，处理 '?' 为 NaN
-            features = [float(f) if f != '?' else np.nan for f in line[:-1]]
-            label = int(line[-1])
-            data.append(features)
-            labels.append(label)
+            if len(line) >= 1 and line[-1].isdigit():
+                line_data = [0 if x == '?' else float(x) for x in line[:-1]]
+                data.append(line_data)
+                labels.append(int(line[-1]))
 
     data = np.array(data)
     labels = np.array(labels)
 
-    # 计算每列的均值，并用均值替换 NaN
-    col_means = np.nanmean(data, axis=0)  # 计算每列的均值
-    for i in range(data.shape[1]):
-        data[:, i] = np.where(np.isnan(data[:, i]), col_means[i], data[:, i])
+    # 检查特征数量
+    num_features = data.shape[1]
 
-    # 标准化数据（均值为 0，标准差为 1）
-    means = np.mean(data, axis=0)
-    stds = np.std(data, axis=0)
-    data = (data - means) / stds
+	    # 特征选择:去掉特征3和特征24~28，确保索引有效
+    indices_to_delete = [2]  # 特征3
+    if num_features > 24:  # 确保特征数量足够
+	        indices_to_delete += list(range(24, 27))  # 特征24到27
 
-    return data, labels
+    # 删除特征
+    data = np.delete(data, indices_to_delete, axis=1)
 
+	    # 将标签二分类化，特征23取值2和3合并为0，其余为1
+    target_variable = np.where((labels == 2) | (labels == 3), 0, 1)
+
+    return data, target_variable
+
+def accuracy(X, y, weights):
+    predictions = predict_logreg(X, weights)
+    return np.mean(predictions == y)
+
+# 其余代码保持不变
 if __name__ == "__main__":
-    # 加载训练集
-    X_train, y_train = load_data("./data/horse-colic.data")
+    # 加载数据
+    X, y = load_data("./data/horse-colic.data")
+    
+    # 拆分训练集和验证集
+    X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)
+
+    # 数据归一化
+    scaler = MinMaxScaler()
+    X_train = scaler.fit_transform(X_train)
+    X_val = scaler.transform(X_val)
+
+    # 训练模型
+    final_weights = train_logreg(X_train, y_train, learning_rate=0.01, iterations=5000, batch_size=64, lambda_reg=0.1)
+
+    # 验证模型
+    val_accuracy = accuracy(X_val, y_val, final_weights)
+    print(f"验证集的准确率: {val_accuracy:.4f}")
 
     # 加载测试集
     X_test, y_test = load_data("./data/horse-colic.test")
+    X_test = scaler.transform(X_test)
 
-    # 训练对率回归模型，加入正则化并增加迭代次数为 5000 次
-    weights = train_logreg(X_train, y_train, learning_rate=0.01, iterations=10000, lambda_reg=0.01)
-
-    # 测试单个数据
-    y_pred = predict_logreg(X_test[11], weights)
-    print(f"预测结果: {y_pred}, 实际结果: {y_test[11]}")
-
-    # 测试整体准确率
-    y_test_pred = np.array([predict_logreg(X_test[i], weights) for i in range(X_test.shape[0])])
-    accuracy = np.mean(y_test_pred == y_test)
-    print(f"测试集准确率: {accuracy:.4f}")
+    # 测试模型
+    final_test_accuracy = accuracy(X_test, y_test, final_weights)
+    print(f"最终测试集的准确率: {final_test_accuracy:.4f}")