diff --git a/chapter_optimization/minibatch-sgd.md b/chapter_optimization/minibatch-sgd.md
index 3afda7763df341c5fe41afa8e4b81cb553f5b204..e876fcbe730ed15e0d80485e34edd5ff1bb04819 100644
--- a/chapter_optimization/minibatch-sgd.md
+++ b/chapter_optimization/minibatch-sgd.md
@@ -165,7 +165,7 @@ train_gluon_ch7('sgd', {'learning_rate': 0.05}, features, labels, 10)
 * 查阅MXNet文档，使用`Trainer`类的`set_learning_rate`函数，令小批量随机梯度下降的学习率每过一个迭代周期减小到原值的1/10。
 
 
-## 扫码直达[讨论区](https://discuss.gluon.ai/t/topic/1877)
+## 扫码直达[讨论区](https://discuss.gluon.ai/t/topic/8013)
 
 ![](../img/qr_minibatch-sgd.svg)