- Approximate the solution, not the problem (no special cases)
- Drive from the problem
- Take the agent’s point of view
- Don’t ask the agent to achieve what it can’t measure
- Don't ask the agent to know what it can't verify
- Set measurable goals for subparts of the agent
- Discriminative models are usually better than generative models
- Work by orthogonal dimensions. Work issue by issue
- Work on ideas, not software
- Experience is the data of AI
- Don’t be impressed by what you don’t understand
- Ideas matter, you should care of that
Interesting:
Model-free RL: TD, E-TD
Model-based RL: feature construction, state constraction
Option abstraction
Super Dyna, Dyna with function approximation
gitbook对数学公式不是很友好，但是代码高亮不错，决定以后在数学公式方面侧重于用印象笔记记录，部分代码方面有gitbook记录。
聊了很多今晚。
Dyna with FA
how to sync with github？
{% hint style="info" %} Drafts --> Merge {% endhint %}