Section: 第10讲神经网络的优化2 | 机器学习（2024年春） | 蕴瑜课堂

Side Block Position

Top Block Position

Section outline

6.5 梯度下降法的学习率
6.6 激活函数
6.7 梯度消失和梯度爆炸问题
6.8 欠拟合、过拟合及其应对方法
- Select activity 10-1 复习 batch 和动量（Momentum NAG）
  
  10-1 复习 batch 和动量（Momentum NAG） File
  
  Students must
  
  Mark as done
- Select activity 10-2 复习自适应学习率(AdaGrad RMSProp) RMSProp+Momentum=Adam
  
  10-2 复习自适应学习率(AdaGrad RMSProp) RMSProp+Momentum=Adam File
  
  Students must
  
  Mark as done
- Select activity 10-3 梯度消失和梯度爆炸BN
  
  10-3 梯度消失和梯度爆炸BN File
  
  Students must
  
  Mark as done
- Select activity 10-4激活函数 dropout
  
  10-4激活函数 dropout File
  
  Students must
  
  Mark as done
- Select activity 课中线上资源批次标准化BN
  
  课中线上资源批次标准化BN File
  
  Students must
  
  Mark as done
- Select activity 课中资源激活函数 dropout
  
  课中资源激活函数 dropout File
  
  Students must
  
  Mark as done

Bottom Block Position