1. Seize the Interval
   1. explore when have time to use the resulting knowledge and exploit when ready to cash in
   2. Exploitation phase means the end of the interval?? 亏则盈，满则溢？
2. Win-stay Lose-shift
   1. Lose shift strategy is rather rough: good options shouldnt be penalized too strongly for being imperfect， 完美主义者来学学。。。。。
   2. doesnt take into account of the interval
   3. endless opportunities to try out, no exact solution
3. The Gittins Index
   1. Assumption: Geometric Discounting
4. Regret and Optimism
   1. pick the option with the highest UCB
      1. dont care the historical record
         1. choose the one with possible highest return in the future
   2. Upper bound confidence
      1. optimism in the face of uncertainty
   3. CONCLUSION
      1. In the long run, optimism is the best prevention for regrets
      2. to assume the best of them, in absence of evidence to the contrary
      3. 问题：如果一旦出现相反的例子该怎么办呢？
5. Baddits Online
   1. A/B testing
      1. optimal way of balance exploration/exploition ??
6. Clinical Trials on Trial
   1. multi-armed bandit 算法在临床试验应用里的争议
7. The Restless World
   1. 如何理解： 对于一个持续不断变化的世界来说，通过进化来调节本能并不一定有助于工业标准化时代？？？
8. 题外延伸：专治选择困难症——bandit算法
9. Explore
   1. explore early on and exploit at a later stage, but sacrifixe the good lay-offs
   2. 在人生的早期阶段，尝试新的冒险的激进事物，
10. Exploit
    1. 听老人言 没错的，是要建立对事物的鉴赏水平和基本观念相似的基础上吧？？
    2. honing social network down to the most meaningful relationships is the rational responce of having less time of enjoy them
11. the anwser is to seize the interval