@makehacklearn Yeah, it’s brute forcing the static evaluator. You can’t brute force the game tree (which is how chess was solved).
-
-
Replying to @Meaningness
@makehacklearn I find it unsurprising & uninteresting that you can brute force the static evaluator; hard to imagine how that could not-work2 replies 0 retweets 0 likes -
Replying to @Meaningness
@Meaningness@makehacklearn What do you mean by brute-forcing the static evaluator? Seems like a contradiction in terms to me.1 reply 0 retweets 0 likes -
Replying to @dfko_0
@dfko_0@makehacklearn Play zillions of games, see what board configurations lead to wins, record that3 replies 0 retweets 0 likes -
Replying to @Meaningness
@Meaningness@makehacklearn Where would you say deep nets lie on continuum from lookup table to perfectly eloquent description of data?1 reply 1 retweet 0 likes -
Replying to @dfko_0
@dfko_0@makehacklearn This depends on hyperparameter settings, problem dimensionality, etc. Can’t make a broad generalization.1 reply 0 retweets 0 likes -
Replying to @Meaningness
@dfko_0@makehacklearn However, the recent successes, and gestalt coming out of Google AI lab, seem to be mostly memorizing masses of data.2 replies 0 retweets 0 likes -
Replying to @Meaningness
@Meaningness@makehacklearn I'm working towards the opposite, smart learning algo, few parameters, very strong regularization1 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.