From caa899933dc9cadcf731317ced2fc4dde5a9cbf0 Mon Sep 17 00:00:00 2001
From: Xionghaoran <719929980@qq.com>
Date: Thu, 31 Oct 2024 15:37:28 +0800
Subject: [PATCH] =?UTF-8?q?=E6=9B=B4=E6=96=B0=20function.py?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 function.py | 30 +++++++++++++++++++++++-------
 1 file changed, 23 insertions(+), 7 deletions(-)

diff --git a/function.py b/function.py
index b6d9a7c..9bd062a 100644
--- a/function.py
+++ b/function.py
@@ -1,6 +1,8 @@
 # function.py
 import numpy as np
 from sklearn.preprocessing import PolynomialFeatures
+from sklearn.feature_selection import SelectFromModel
+from sklearn.linear_model import LogisticRegression
 
 """
 实验说明:
@@ -21,8 +23,8 @@ from sklearn.preprocessing import PolynomialFeatures
 """
 请在这里填写学号和姓名，不填写即成绩无效。
 """
-id = "202210423213"
-name = "熊浩燃"
+id = "202210423xxx"
+name = "test_name"
 
 def sigmoid(z):
     return 1 / (1 + np.exp(-z))
@@ -36,9 +38,19 @@ def add_polynomial_features(X, degree=2):
     poly = PolynomialFeatures(degree)
     return poly.fit_transform(X)
 
+def select_features(X, y, penalty='l1', C=1.0):
+    """
+    使用L1正则化选择重要特征。
+    """
+    model = LogisticRegression(penalty=penalty, solver='liblinear', C=C)
+    model.fit(X, y)
+    selector = SelectFromModel(model, prefit=True)
+    X_selected = selector.transform(X)
+    return X_selected, selector
+
 def train_logreg(X, y, learning_rate=0.0001, iterations=5000, batch_size=32, lambda_=0.1, degree=2):
     """
-    对率回归训练函数，包含批量梯度下降和L2正则化，并增加多项式特征。
+    对率回归训练函数，包含批量梯度下降和L1正则化，并增加多项式特征。
 
     Args:
     - X (numpy array): 特征数据。
@@ -51,9 +63,11 @@ def train_logreg(X, y, learning_rate=0.0001, iterations=5000, batch_size=32, lam
 
     Returns:
     - numpy array: 训练得到的权重。
+    - selector: 特征选择器，用于预测时筛选相同特征。
     """
     X = standardize(X)  # 标准化数据
     X = add_polynomial_features(X, degree=degree)  # 增加多项式特征
+    X, selector = select_features(X, y)  # 特征选择
     m, n = X.shape
     weights = np.random.rand(n) * 0.01  # 初始化权重为小的随机数
     
@@ -72,15 +86,16 @@ def train_logreg(X, y, learning_rate=0.0001, iterations=5000, batch_size=32, lam
         gradient = np.dot(X_batch.T, (h - y_batch)) / effective_batch_size + lambda_ * weights
         weights -= learning_rate * gradient
     
-    return weights
+    return weights, selector
 
-def predict_logreg(inX, weights, degree=2):
+def predict_logreg(inX, weights, selector, degree=2):
     """
     对率回归预测函数。
 
     Args:
     - inX (numpy array): 输入待分类的样本。
     - weights (numpy array): 训练得到的权重。
+    - selector: 特征选择器，用于筛选特征。
     - degree (int): 多项式特征的次数。
 
     Returns:
@@ -88,6 +103,7 @@ def predict_logreg(inX, weights, degree=2):
     """
     inX = standardize(inX.reshape(1, -1))  # 标准化输入
     inX = add_polynomial_features(inX, degree=degree)  # 增加多项式特征
+    inX = selector.transform(inX)  # 应用特征选择，确保预测时与训练时使用相同的特征
     probability = sigmoid(np.dot(inX, weights))
     return 1 if probability >= 0.5 else 0
 
@@ -129,8 +145,8 @@ if __name__ == "__main__":
     X_test, y_test = load_data("./data/horse-colic.test")
 
     # 训练对率回归模型
-    weights = train_logreg(X_train, y_train, learning_rate=0.0001, iterations=5000, batch_size=32, lambda_=0.1, degree=2)
+    weights, selector = train_logreg(X_train, y_train, learning_rate=0.0001, iterations=5000, batch_size=32, lambda_=0.1, degree=2)
 
     # 测试单个数据
-    y_pred = predict_logreg(X_test[11], weights, degree=2)
+    y_pred = predict_logreg(X_test[11], weights, selector, degree=2)
     print(f"预测结果: {y_pred}, 实际结果: {y_test[11]}")