3 ms·
They should incorporate ML options too. They have multivariate testing + goal/conversion measurement, they should have multi-armed bandit ML methods as an optio
by cphoover 10y ago
They should incorporate ML options too. They have multivariate testing + goal/conversion measurement, they should have multi-armed bandit ML methods as an option (epsilon-greedy, bayesian bandit etc).
- pkaeding 10y agoAs a matter of fact, we have experimented with multi-armed bandit 'auto-pilot' to roll out a new feature while watching A/B test results. I'd like to see this make it to production as well; it was one of those things that sounded like a fun idea, but we weren't sure if actual customers would use it :). nb: I work for LaunchDarkly, if it wasn't obvious from my comment.