From d9385b0323eea575212a3f0c16433a3d3a79a563 Mon Sep 17 00:00:00 2001
From: cy <939601402@qq.com>
Date: Thu, 24 Oct 2024 17:21:10 +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 | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/function.py b/function.py
index ed75430..daa5782 100644
--- a/function.py
+++ b/function.py
@@ -28,7 +28,7 @@ def sigmoid(z):
     return 1.0 / (1 + np.exp(-z))
 
 # 这里写自己的代码实现
-def train_logreg(X, y, learning_rate=0.01, iterations=1000, batch_size=32):
+def train_logreg(X, y, learning_rate=0.01, iterations=1000):
     """
     对率回归训练函数。
 
@@ -42,15 +42,15 @@ def train_logreg(X, y, learning_rate=0.01, iterations=1000, batch_size=32):
     """
     m, n = X.shape
     weights = np.zeros(n)
+    # 用于Adagrad的累积平方梯度
+    r = np.zeros(n)
+    epsilon = 1e-8
     for i in range(iterations):
-        for j in range(0, m, batch_size):
-            end = min(j + batch_size, m)
-            X_batch = X[j:end]
-            y_batch = y[j:end]
-            z = np.dot(X_batch, weights)
-            h = sigmoid(z)
-            gradient = np.dot(X_batch.T, (h - y_batch)) / len(X_batch)
-            weights -= learning_rate * gradient
+        z = np.dot(X, weights)
+        h = sigmoid(z)
+        gradient = np.dot(X.T, (h - y)) / m
+        r += gradient ** 2
+        weights -= learning_rate * gradient / (np.sqrt(r) + epsilon)
     return weights
 
 def predict_logreg(inX, weights):