Self-Optimizing A/B Tests

Re:Self-Optimizing A/B Tests by isoprophlex on Tuesday October 12, 2021 @11:11AM (#28840705)

One trick I used myself in a bayesian bandit-esque approach (thompson sampling from some distribution, eg. a Beta distribution) was to implement some "forgetting rate" on the parameters of the distribution.

[ Parent | on HN ]

Re:Self-Optimizing A/B Tests by chanind on Tuesday October 12, 2021 @12:06PM (#28841405)

I updated the post to note Thompson Sampling and added a note about using only recent data as a potential improvement and linked to this comment. Thanks for this note!

[ Parent | on HN ]