更新 function.py

function.py

@@ -1,10 +1,7 @@
 # function.py
 import numpy as np
 import pandas as pd
-from sklearn.model_selection import train_test_split, GridSearchCV
 from sklearn.preprocessing import StandardScaler
-from sklearn.neighbors import KNeighborsClassifier
-from sklearn.metrics import accuracy_score
 
 """
 挑战说明:
@@ -46,6 +43,40 @@ name = "熊浩燃"
 
 
 # 这里写自己的代码实现
+def classify0(inX, dataSet, labels, k=3):
+    """
+    KNN 分类算法，使用马氏距离。
+
+    Args:
+    - inX (numpy array): 输入待分类的样本。
+    - dataSet (numpy array): 训练数据集。
+    - labels (list or numpy array): 训练集对应的标签。
+    - k (int): kNN 算法参数，默认值为 3。
+
+    Returns:
+    - int: 预测的分类标签。
+    """
+    # 标准化数据
+    scaler = StandardScaler()
+    dataSet = scaler.fit_transform(dataSet)
+    inX = scaler.transform([inX])[0]
+    
+    # 计算协方差矩阵的逆
+    cov_matrix = np.cov(dataSet, rowvar=False)
+    inv_cov_matrix = np.linalg.inv(cov_matrix)
+    
+    # 计算马氏距离
+    diff = dataSet - inX
+    distances = np.sqrt(np.sum(diff @ inv_cov_matrix * diff, axis=1))
+    
+    # 按距离排序并返回 k 个最近的标签
+    sorted_indices = distances.argsort()
+    nearest_labels = [labels[i] for i in sorted_indices[:k]]
+    
+    # 找出出现次数最多的标签
+    label_counts = np.bincount(nearest_labels)
+    return label_counts.argmax()
+
 def load_data(file_path):
     """
     加载 TSV 格式数据文件。
@@ -57,41 +88,38 @@ def load_data(file_path):
 
 def find_best_k(X_train, y_train):
     """
-    使用交叉验证找到最佳的 k 值。
+    寻找最佳 k 值。
     """
-    param_grid = {'n_neighbors': range(1, 21), 'weights': ['uniform', 'distance']}
-    knn = KNeighborsClassifier()
-    grid_search = GridSearchCV(knn, param_grid, cv=5)
-    grid_search.fit(X_train, y_train)
-    return grid_search.best_params_
+    best_k = 1
+    best_accuracy = 0
+
+    for k in range(1, 21):
+        # 使用 classify0 进行预测
+        y_pred = [classify0(x, X_train, y_train, k) for x in X_train]
+        accuracy = np.mean(y_pred == y_train)
+
+        if accuracy > best_accuracy:
+            best_accuracy = accuracy
+            best_k = k
+
+    return best_k, best_accuracy
 
 if __name__ == "__main__":
-    # 加载数据
     train_file = "datingTestSet2.tsv"
     X, y = load_data(train_file)
 
-    # 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)
-
     # 标准化特征
     scaler = StandardScaler()
-    X_train = scaler.fit_transform(X_train)
-    X_test = scaler.transform(X_test)
+    X = scaler.fit_transform(X)
 
     # 寻找最佳 k 值
-    best_params = find_best_k(X_train, y_train)
-    best_k = best_params['n_neighbors']
-    best_weights = best_params['weights']
-
-    # 使用最佳参数训练模型并评估
-    knn = KNeighborsClassifier(n_neighbors=best_k, weights=best_weights)
-    knn.fit(X_train, y_train)
-    y_pred = knn.predict(X_test)
-    
-    # 输出准确率
-    accuracy = accuracy_score(y_test, y_pred)
-    print(f"最佳 k 值: {best_k}, 权重: {best_weights}, 准确率: {accuracy:.2f}")
-
-
+    best_k, best_accuracy = find_best_k(X, y)
 
+    # 输出最佳 k 值和对应的准确率
+    print(f"最佳 k 值: {best_k}, 准确率: {best_accuracy:.2f}")
 
+    # 进行预测
+    sample = np.array([65915, 3.868359, 0.128078])
+    sample = scaler.transform([sample])[0]  # 标准化样本
+    predicted_label = classify0(sample, X, y, best_k)
+    print(f"预测结果: {predicted_label}")