更新 function.py

function.py

@@ -6,7 +6,7 @@ name = "吴杨洋"
 def sigmoid(z):
     return 1 / (1 + np.exp(-np.clip(z, -500, 500)))
 
-def train_logreg(X, y, numIter=1000, learning_rate=0.01, lambda_reg=0.01):
+def train_logreg(X, y, numIter=1 + 1000, learning_rate=0.01, lambda_reg=0.01):
     m, n = X.shape
     weights = np.zeros(n)
     for j in range(numIter):
@@ -17,7 +17,7 @@ def train_logreg(X, y, numIter=1000, learning_rate=0.01, lambda_reg=0.01):
             gradient = X[i] * error
             reg_term = lambda_reg * weights
             gradient += reg_term
-            weights += learning_rate * gradient
+            weights -= learning_rate * gradient
     return weights
 
 def predict_logreg(inX, weights):
@@ -32,21 +32,20 @@ def load_data(filepath):
     y = []
     for data in fr:
         data = data.strip()
-        data = data.split(' ')
+        data = data.split()
+        if len(data) < 25:  # 确保每行数据有足够的元素
+            continue
+        y.append(int(data[22]) > 1)  # 假设标签是 0 或 1
         data_25 = data[24]
-        y.append(data[222])
-        del data[-1]
-        del data[26]
-        del data[25]
-        del data[24]
-        del data[222]
-        del data[2]
-        for i in data_25:
+        # 删除不需要的元素
+        del data[0], data[1], data[2], data[22], data[24], data[25]
+        for i in data_25.split(','):
             data.append(i)
-        data = [0 if data_elem == '?' else float(data_elem) for data_elem in data]
+        # 将 '?' 替换为 0
+        data = [0 if x == '?' else float(x) for x in data]
         data_list.append(data)
     data_number = len(y)
-    y_train = np.array([1 if y[j] == '2' or y[j] == '3' else 0 for j in range(data_number)])
+    y_train = np.array(y)
     X = np.array(data_list)
     return X, y_train
 
@@ -62,22 +61,11 @@ def main():
     X_train, mean, std = standardize(X_train)
     X_test = (X_test - mean) / std
 
-    # 尝试不同的学习率和正则化参数
-    learning_rates = [0.01, 0.001, 0.0001]
-    lambda_regs = [0.1, 0.01, 0.001]
+    weights = train_logreg(X_train, y_train, numIter=1000, learning_rate=0.01, lambda_reg=0.1)
 
-    best_accuracy = 0
-    best_weights = None
-    for lr in learning_rates:
-        for reg in lambda_regs:
-            weights = train_logreg(X_train, y_train, numIter=1000, learning_rate=lr, lambda_reg=reg)
-            y_pred_test = [predict_logreg(x, weights) for x in X_test]
-            accuracy_test = np.mean(y_pred_test == y_test)
-            if accuracy_test > best_accuracy:
-                best_accuracy = accuracy_test
-                best_weights = weights
-
-    print(f"Best Test Accuracy: {best_accuracy:.2f}")
+    y_pred_test = [predict_logreg(x, weights) for x in X_test]
+    accuracy_test = np.mean(y_pred_test == y_test)
+    print(f"Test Accuracy: {accuracy_test:.2f}")
 
 if __name__ == "__main__":
     main()