Hi, I'm getting a db connection error when trying to view the article so I'm not sure what it's all about yet.
I'm aware of your work with racket (read your paper from a few years ago about deploying the web server in a uni environment).
Is this something I can plug into a racket web app? If it is I'd be interested as I am building web apps with racket. I'm not in a position where I can actually deploy something yet but I'll want to build in statistics collection when I am a little closer.
Edit: OK, I read your article via the cached link another user posted. I'll have a look at your linked papers etc when I get a chance but it definitely looks interesting. Your post wasn't clear on what form this beta will take - is it a commercial venture to compete with optimizely et al or one of your open source projects I could require into my own app and deploy myself?
We're making a commercial system. Implementing a bandit algorithm yourself is quite straightforward. You could implement, say, UCB1 in a day. I might open source some of that code if I get the time. Packaging that algorithm into a usable and scalable system (unlike our blog ;-) is a lot more work.
PS: I've installed a cache so the blog should stay up now.
I'm aware of your work with racket (read your paper from a few years ago about deploying the web server in a uni environment).
Is this something I can plug into a racket web app? If it is I'd be interested as I am building web apps with racket. I'm not in a position where I can actually deploy something yet but I'll want to build in statistics collection when I am a little closer.
Edit: OK, I read your article via the cached link another user posted. I'll have a look at your linked papers etc when I get a chance but it definitely looks interesting. Your post wasn't clear on what form this beta will take - is it a commercial venture to compete with optimizely et al or one of your open source projects I could require into my own app and deploy myself?