【Slides】一天搞懂深度学习 by 台湾大学电机工程学助理教授李宏毅

Backto DL Materials

链接: https://pan.baidu.com/s/1SzfuZtyFmNY2w2aMuM2xaA 提取码: 3xfe

Lecture I: Introduction of Deep Learning

DL 统共分三步，

• Step 1: Desgin a Network Struture (a set of functions with params $\theta$, linear/non-linear)
• Step 2: Set Learning Target (loss function $L$)
• Step 3: Dive into learning! (pick the best group of $\theta$ that minimize $L$)
  
  DL 胜任的工作:
  

Human Brains

AI Neural Network

A typical Fully Connected Network

Loss

Total Loss

• Gradient Descent
  

• Backpropagation is an efficient way to compute $\partial L / \partial w$
  

Lecture II: Tips for Training Deep Neural Network

Lecture III: Variants of Neural Network

Lecture IV: Next Wave