So the future of DL looks like genetic programming? What do you think DL brings to the table that will make it more successful?
Some of my thoughts on the future of deep learning: https://blog.keras.io/the-future-of-deep-learning.html … - This is the follow-up to yesterday's post on the limits of DL.
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
for hyperparameter tuning,
@SigOpt is working on implementing hyperparameter optimization (i.e. less involvement from humans)@DrScottClarkThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
You are spot on. We share the same vision. Working on my Cortex Adsembler language in that direction.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Especially liked the anthropomorphic part in first one, agree with search methods (gp). novelty, MAP-elites + sgd will be relevant. Thanks!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
One thing we should do better in future anyway is making ML more accessible and more comprehensible for the public. Transparency!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Good points. To achieve "extreme generalization" I think context where each learning is done will be key.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
You are right DL is fake
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
"that's a pure fantasy, coming from a long series of profound misunderstandings of both intelligence and technology".So no bestselling books
-
in 30 years or so from now? https://www.technologyreview.com/s/607970/ . While many indeed don't have a clue about the essence of DL,such claims are mostly hype
End of conversation
New conversation -
-
-
Very interesting. Thanks!
Thanks. 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.