更新 function.py

function.py

@@ -23,8 +23,8 @@ import numpy as np
 姓名请直接**填写中文**
 例如: name = "张三"
 """
-id = "202210423xxx"
+id = "202210423130"
-name = "test_name"
+name = "董秋"
 
 
 # 这里写自己的代码实现
@@ -40,7 +40,22 @@ def train_svm(X, y, other_params="你自己设置一些其它参数"):
     Returns:
     - numpy array: 训练得到的权重。
     """
+    n_samples, n_features = X.shape
+    weights = np.zeros(n_features)  # 初始化权重
+    bias = 0  # 初始化偏置
     
+    for _ in range(iterations):
+        for idx, x_i in enumerate(X):
+            condition = y[idx] * (np.dot(x_i, weights) + bias) >= 1
+            if condition:
+                # 更新公式，不需要正则化项
+                weights -= learning_rate * (2 * lambda_param * weights)
+            else:
+                # 更新公式，包括误分类损失
+                weights -= learning_rate * (2 * lambda_param * weights - np.dot(x_i, y[idx]))
+                bias -= learning_rate * y[idx]
+    
+    return np.append(weights, bias)  # 将偏置合并到权重中
 
 def predict_svm(inX, weights):
     """
@@ -53,7 +68,10 @@ def predict_svm(inX, weights):
     Returns:
     - int: 预测的分类标签。
     """
-
+    w = weights[:-1]  # 分离权重和偏置
+    b = weights[-1]
+    linear_output = np.dot(inX, w) + b
+    return 1 if linear_output >= 0 else -1
 
 # 这里也写自己的代码实现
 def load_data(filepath):
@@ -67,7 +85,16 @@ def load_data(filepath):
     - np.array: 特征数据。
     - np.array: 标签数据。
     """
-
+    data = []
+    with open(filepath, 'r') as file:
+        for line in file:
+            if "?" not in line:  # 去除包含缺失值的数据行
+                data.append([float(x) for x in line.strip().split()])
+    data = np.array(data)
+    X = data[:, :-1]  # 所有列，除最后一列
+    y = data[:, -1]   # 最后一列
+    y = np.where(y == 1, 1, -1)  # 将标签转换为 +1 和 -1
+    return X, y
 
 if __name__ == "__main__":
     # 以下代码可用于测试，可自行随意修改