@avibryant @tixxit did you see this paper about explaining black box models? http://www.blackboxworkshop.org/pdf/Turner2015_MES.pdf …
-
-
Replying to @avibryant
@avibryant@b0rk I like the idea of standardizing on a general explanation output (and we can cheat really easily with RFs).1 reply 0 retweets 2 likes -
Replying to @tixxit
@tixxit@avibryant the definition of a scoring function seems super useful & I'm impressed that they got an interpretable image result1 reply 0 retweets 0 likes -
Replying to @b0rk
@b0rk@avibryant If we have the "input density" though, I do kind of wonder what simply building a decision tree on a sample would yield3 replies 0 retweets 0 likes -
Replying to @tixxit
@tixxit@avibryant it seems nontrivial to do calculations using the input density (but I think it's possible for a random forest)1 reply 0 retweets 0 likes -
Replying to @avibryant
@b0rk@tixxit and so we can at least pick the best explanation from the set of prefixes of paths from root to the input's leaf.1 reply 0 retweets 0 likes
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.