[don't bother with reinforcement learning] video notes (I) 3. Why use reinforcement learning?

【 Don't bother with reinforcement learning videos 】 The notebook

The first 3 section Why use reinforcement learning ？

Reinforcement learning is a big branch of machine learning , With the development of recent years , Reinforcement learning is also often combined with deep learning . in general , Reinforcement learning is to make your computer From scratch To learn , No need for any “ supervise （Um…… Basically, self-study , From little white to big guy ） Reference resources ” To learn how to choose actions , How to act to get Higher returns . I said before , Reinforcement learning has “ Score orientation ”.
Here are some small examples of reinforcement learning , The environment is, for example, maze , Where is the starting point , There is a wall , Where is the end （ These can be called States ）, And the probability of state transition , Visualization and so on , Actions such as southeast, northwest 、 Up, down, left, right, etc . The computer keeps trying （ Early stage is like mental retardation ）, And then constantly sum up the experience （ Update strategy ）, The process of finally obtaining the optimal strategy .
youku Reinforcement learning simulation video

• Robots go through mazes
  
• Inverted pendulum
  
• Car climbing
  Last one ：【 Don't bother to strengthen learning 】 Video notes （ One ）2. Summary of reinforcement learning methods
  Next ：【 Don't bother to strengthen learning 】 Video notes （ Two ）1. What is? Q Learning