更新 function.py

function.py

@@ -1,8 +1,7 @@
 # function.py
 import numpy as np
 from sklearn.preprocessing import StandardScaler
-from itertools import product
+from sklearn.model_selection import train_test_split, KFold
-from sklearn.ensemble import RandomForestClassifier
 
 """
 实验说明:
@@ -27,92 +26,101 @@ id = "202210423117"
 name = "程悦"
 
 def sigmoid(z):
-    return 1.0 / (1 + np.exp(-z))
+    """Sigmoid 激活函数。"""
+    return 1 / (1 + np.exp(-z))
 
-
+def train_logreg(X, y, learning_rate=0.1, iterations=2000, lambda_reg=0.01, batch_size=32):
-def train_logreg(X, y, learning_rate=0.05, iterations=1500, batch_size=64, patience=100, decay_rate=0.95):
+    """逻辑回归训练函数，随机初始化权重，并在每次迭代前打乱数据。"""
     m, n = X.shape
-    weights = np.random.randn(n) * 0.01
+    weights = np.random.randn(n) * 0.01  # 随机初始化权重，取较小值
-    best_weights = weights.copy()
-    best_accuracy = 0
-    wait = 0
-    decay_step = iterations // 10  # 每10%的迭代衰减一次学习率
 
     for i in range(iterations):
+        # 每次迭代打乱数据
         indices = np.random.permutation(m)
         X_shuffled = X[indices]
         y_shuffled = y[indices]
 
-        for j in range(0, m, batch_size):
+        for j in range(0, m, batch_size):  # 按照 batch size 迭代
             X_batch = X_shuffled[j:j + batch_size]
             y_batch = y_shuffled[j:j + batch_size]
+
             z = np.dot(X_batch, weights)
-            h = sigmoid(z)
+            predictions = sigmoid(z)
-            gradient = np.dot(X_batch.T, (h - y_batch)) / len(y_batch)
+            gradient = np.dot(X_batch.T, (predictions - y_batch)) / X_batch.shape[0]
+
+            # L2正则化
+            gradient += (lambda_reg / m) * weights
+
+            # 更新权重
             weights -= learning_rate * gradient
 
-            # 计算训练集准确率
+        # 动态调整学习率，逐渐减小
-        train_accuracy = accuracy(X, y, weights)
+        if i % 100 == 0:
-        print(f"Iteration {i + 1}, Train Accuracy: {train_accuracy:.4f}")
+            learning_rate *= 0.95
-
-        if train_accuracy > best_accuracy:
-            best_accuracy = train_accuracy
-            best_weights = weights.copy()
-            wait = 0
-        else:
-            wait += 1
-
-        if i % decay_step == 0 and i > 0:
-            learning_rate *= decay_rate
-            print(f"Learning rate decayed to: {learning_rate:.6f}")
-
-        if wait >= patience:
-            print("Early stopping!")
-            break
-
-    return best_weights
 
+    return weights
 
 def predict_logreg(X, weights):
+    """逻辑回归预测函数。"""
     z = np.dot(X, weights)
-    h = sigmoid(z)
+    prediction_prob = sigmoid(z)
-    return np.where(h > 0.5, 1, 0)
+    return np.where(prediction_prob >= 0.5, 1, 0)
-
 
 def load_data(filepath):
-    data = []
+    data, labels = [], []
-    labels = []
     with open(filepath, 'r') as f:
-        for line in f:
+        for line in f.readlines():
-            parts = line.strip().split()
+            line = line.strip().split()
-            if len(parts) >= 2 and parts[-1].isdigit():
+            if len(line) >= 1 and line[-1].isdigit():
-                line_data = [0 if x == '?' else float(x) for x in parts[:-1]]
+                line_data = [1] + [0 if x == '?' else float(x) for x in line[:-1]]
-                label = int(parts[-1])
-                # 将标签映射到0和1（假设原始标签是1, 2, 3，我们将其映射为0和1）
-                target_label = 0 if label in [2, 3] else 1
                 data.append(line_data)
-                labels.append(target_label)
+                labels.append(int(line[-1]))
 
     data = np.array(data)
     labels = np.array(labels)
-    return data, labels
 
+    target_variable = labels
+    target_variable[(target_variable == 2) | (target_variable == 3)] = 0
+
+    data = np.delete(data, [2] + list(range(23, 28)), axis=1)
+
+    scaler = StandardScaler()
+    data = scaler.fit_transform(data)
+
+    return data, target_variable
 
 def accuracy(X, y, weights):
+    """计算模型准确率。"""
     predictions = predict_logreg(X, weights)
     return np.mean(predictions == y)
 
-
 if __name__ == "__main__":
-    X_train, y_train = load_data("./horse-colic.data")
+    # 加载数据集
-    X_test, y_test = load_data("./horse-colic.test")
+    X, y = load_data("./horse-colic.data")
 
-    # 标准化数据
+    # 使用 train_test_split 打乱并随机抽取数据
-    scaler = StandardScaler()
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
-    X_train = scaler.fit_transform(X_train)
-    X_test = scaler.transform(X_test)
 
-    final_weights = train_logreg(X_train, y_train, learning_rate=0.001, iterations=1500, batch_size=64, patience=200)
+    # K折交叉验证
+    k = 5
+    kf = KFold(n_splits=k, shuffle=True, random_state=42)
+    accuracies = []
 
-    final_accuracy = accuracy(X_test, y_test, final_weights)
+    for train_index, val_index in kf.split(X_train):
-    print(f"最终测试集的准确率: {final_accuracy:.4f}")
+        X_train_kf, X_val_kf = X_train[train_index], X_train[val_index]
+        y_train_kf, y_val_kf = y_train[train_index], y_train[val_index]
+
+        # 训练逻辑回归模型，设置 batch size
+        weights = train_logreg(X_train_kf, y_train_kf, learning_rate=0.1, iterations=2000, lambda_reg=0.01, batch_size=32)
+
+        # 验证模型
+        val_accuracy = accuracy(X_val_kf, y_val_kf, weights)
+        accuracies.append(val_accuracy)
+
+    # 输出交叉验证的准确率和平均准确率
+    print(f"K折交叉验证得分: {accuracies}")
+    print(f"平均准确率: {np.mean(accuracies):.4f}")
+
+    # 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]}")