训练神经网络(上)——批量归一化
Batch Normalization
![在这里插入图片描述](https://img-blog.csdnimg.cn/20200908221001328.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQyNzE4ODg3,size_16,color_FFFFFF,t_70#pic_center)
Babysitting the learning Process
Step1: Preprocess the data
Step2: Choose the architecture of network
Step3: Double check that the loss is reasonable
Summary
![在这里插入图片描述](https://img-blog.csdnimg.cn/20200908224156230.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQyNzE4ODg3,size_16,color_FFFFFF,t_70#pic_center)
posted @
2020-09-08 22:43
JHadoop
阅读(
95)
评论()
编辑
收藏
举报