Bandit Algorithms Book [pdf] • Hacker News Board

Re: Bandit Algorithms Book [pdf]

#11

Post by bochi » Tue, Jul 31, 2018, 1:10 PM UTC

tomkat0789 wrote: post #8

Never heard of bandit algorithms before! Or if I did I didn't recognize it as something different from probability. What have people around here used them for?

Bandits are probably one of the most underrated machine learning algorithms. One possible application is recommendation systems. Shameless self promotion. I wrote an article about it: https://towardsdatascience.com/how-not-to-sort-by-popularity...

Re: Bandit Algorithms Book [pdf]

#12

Post by srean » Tue, Jul 31, 2018, 2:34 PM UTC

HugoDaniel wrote: post #5

Is this the book that is going to make me a poker master player ?

If you play long enough it will make you regret less

Re: Bandit Algorithms Book [pdf]

#13

Post by ur-whale » Tue, Jul 31, 2018, 3:12 PM UTC

zdkl wrote: post #9

tomkat0789 wrote: post #8

Never heard of bandit algorithms before! Or if I did I didn't recognize it as something different from probability. What have people around here used them for?

This rust project, to manage the number of threads in a monero miner afair. https://github.com/Ragnaroek/mithril

Doesn't alphago use some form of Bandit algorithm in their MonteCarlo code?

Re: Bandit Algorithms Book [pdf]

#14

Post by magoghm » Tue, Jul 31, 2018, 3:26 PM UTC

ur-whale wrote: post #13

zdkl wrote: post #9

Earlier quoted context omitted.

This rust project, to manage the number of threads in a monero miner afair. https://github.com/Ragnaroek/mithril

Doesn't alphago use some form of Bandit algorithm in their MonteCarlo code?

I believe that Monte Carlo Tree Search, used in AlphaGo, does work using bandit algorithms. On top of that AlphaGo uses Reinforcement Learning, which also uses bandit algorithms (in Sutton & Barto's book, "Reinforcement Learning: An Introduction", all of chapter 2 is about multi-armed bandits).

Re: Bandit Algorithms Book [pdf]

#15

Post by mlechha » Tue, Jul 31, 2018, 3:46 PM UTC

tomkat0789 wrote: post #8

Never heard of bandit algorithms before! Or if I did I didn't recognize it as something different from probability. What have people around here used them for?

They're probably the most fundamental kind of reinforcement learning algorithms. Understanding bandit algorithms is crucial to developing a good understanding of RL.

Re: Bandit Algorithms Book [pdf]

#16

Post by pronoiac » Tue, Jul 31, 2018, 4:46 PM UTC

This came up a couple of days ago: https://news.ycombinator.com/item?id=17637683

Re: Bandit Algorithms Book [pdf]

#17

Post by joshuamorton » Tue, Jul 31, 2018, 6:54 PM UTC

It always makes me sad that Thompson Sampling isn't (or at least doesn't appear to be) mentioned alongside things like UCB1. Its theoretically optimal, and relatively easy to grok, and not significantly more difficult to implement.