I like David McAllester's thoughts on this (informal, but interesting)https://machinethoughts.wordpress.com/2016/06/20/cognitive-architectures/ …
Stacked affine transforms are not a very good substrate for most machine learning models. Any sufficiently complicated deep learning model contains ad-hoc, bug-ridden, slow approximations of if-else statements (and other control flow).
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
"slow approximations of if-else statements", what about classical decision trees?
-
This Tweet is unavailable.
- Show replies
New conversation -
-
-
Can those structures be precisely localised?
-
good question - check out 1980s deep science
#NeuralNetworks research, eg.https://link.springer.com/article/10.1007/PL00022037 …
End of conversation
New conversation -
-
-
bad news for applications in
#legaltechThanks. Twitter will use this to make your timeline better. UndoUndo
-
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.