Introduction to Thompson Sampling: The Bernoulli Bandit (2017)

My favorite resource on Thompson Sampling is &lt;<a href="https:&#x2F;&#x2F;everyday-data-science.tigyog.app&#x2F;a-b-testing" rel="nofollow">https:&#x2F;&#x2F;everyday-data-science.tigyog.app&#x2F;a-b-testing</a>&gt;.<p>After learning about it, I went on to replace the UCT formula in MCTS with it and the results were... not much better, actually. But it made me understand both a little better.

Thompson Sampling, a.k.a. Bayesian Bandits, is a powerful method for runtime performance optimization. We use it in ClickHouse to optimize compression and to choose between different instruction sets: <a href="https:&#x2F;&#x2F;clickhouse.com&#x2F;blog&#x2F;lz4-compression-in-clickhouse" rel="nofollow">https:&#x2F;&#x2F;clickhouse.com&#x2F;blog&#x2F;lz4-compression-in-clickhouse</a>

This is great. I remember finding another really good resource on the Bernoulli bandit that was interactive. Putting feelers out there to see if anyone knows what I’m talking about off the top of their heads.

I built a contextual bandit combining XGBoost with Thompson Sampling you can check out at <a href="https:&#x2F;&#x2F;improve.ai" rel="nofollow">https:&#x2F;&#x2F;improve.ai</a>

if you have an NN that is probabilistic, how do you update the prior after sampling from the posterior?

Beautifully composed article. Looking forward to trying this out.