1

function.py

@@ -42,6 +42,7 @@ def load_data(filepath):
         del data[26]
         del data[25]
         del data[24]
+        del data[23]
         del data[22]
         del data[2]
         for i in data_25:
@@ -57,7 +58,7 @@ def load_data(filepath):
 
     return X_train, y_train
 
-def train_logreg(X, y, numIter=1000, learning_rate=0.06, batch_size=128, lambda_reg=0.01):
+def train_logreg(X, y, numIter=1500, learning_rate=0.05, batch_size=64, lambda_reg=0.01):
     m, n = X.shape
     weights = np.zeros(n)
     for j in range(numIter):
@@ -70,9 +71,10 @@ def train_logreg(X, y, numIter=1000, learning_rate=0.06, batch_size=128, lambda_
             h = 1 / (1 + np.exp(-z))
             error = batch_y - h
             gradient = np.dot(batch_X.T, error) / len(batch_y)
-            reg_term = 2 * lambda_reg * weights
-            gradient -= reg_term / len(batch_y)
-            weights += learning_rate * gradient
+            # L1正则化项
+            l1_reg_term = np.sign(weights)
+            # 更新权重，加入L1正则化
+            weights -= learning_rate * (gradient + lambda_reg * l1_reg_term)
     return weights
 
 def predict_logreg(inX, weights):