更新 function.py

function.py

@@ -40,12 +40,18 @@ name = "陈垣豪"
 """
 
 def classify0(inX, dataSet, labels, k=3):
+    # Normalize the dataset
+    dataSet_mean = np.mean(dataSet, axis=0)
+    dataSet_std = np.std(dataSet, axis=0)
+    normalized_dataSet = (dataSet - dataSet_mean) / dataSet_std
+    normalized_inX = (inX - dataSet_mean) / dataSet_std
+    
     # 计算协方差矩阵并求其逆矩阵
-    cov_matrix = np.cov(dataSet, rowvar=False)
+    cov_matrix = np.cov(normalized_dataSet, rowvar=False)
     inv_cov_matrix = np.linalg.inv(cov_matrix)
     
     # 计算马氏距离
-    diff = dataSet - inX
+    diff = normalized_dataSet - normalized_inX
     distances = np.sqrt(np.sum(diff @ inv_cov_matrix * diff, axis=1))
     
     # 获取距离最小的 k 个样本的索引
@@ -56,11 +62,13 @@ def classify0(inX, dataSet, labels, k=3):
     
     # 统计 k 个最近邻中每个类别的出现次数
     label_count = {}
-    for label in nearest_labels:
+    weights = 1 / (distances[sorted_indices] + 1e-10)  # Avoid division by zero
+    
+    for i, label in enumerate(nearest_labels):
         if label in label_count:
-            label_count[label] += 1
+            label_count[label] += weights[i]
         else:
-            label_count[label] = 1
+            label_count[label] = weights[i]
     
     # 找到出现次数最多的标签
     predicted_label = max(label_count, key=label_count.get)