These are a few of my faaaavorite things!
Ah, multi-armed bandits. The old "explore vs exploit" tradeoff. Or as I like to call it, "I wish I spent more time reading through that Distributed Systems book instead of browsing Reddit, but then again I found that book on Reddit to begin with."
Awesome stuff! Making any progress whatsoever in the quest to get ML algos to be able to generalize from experience to new contexts is awesome.
(maybe I should do a longer post later on this topic?)
Hackers and Slackers Newsletter
Join the newsletter to receive the latest updates in your inbox.