Many people have tried to defend pure DRL w Nature Machine Intelligence article that actually is a hybrid model; the below youtube pointer IS to a pure convnet - but read fine print: “network works decently well for any position less than 6 moves away from solved”
#symbolphobiahttps://twitter.com/AlexRoseGames/status/1186571935611850752 …
-
I personally agree that MCTS is a symbolic planning algorithm, even when you plan in domains that have a somewhat opaque forward model (e.g. ALE or GVGAI). But I suppose most people think of symbolic methods as involving a language with predicates, e.g. first-order logic.
i am most concerned with quantified statements, which seem to me to demand “rules” or operations over variables. can you build MCTS without them? (nobody does in practice AFAIK)
-
-
It's actually an interesting question whether you could build MCTS in a completely "neural" way. Maybe you could contort a network to represent the algorithm using recurrent nets and attention layers somehow. But even if you could represent the algorithm, how could you learn it?
-
This is a recent work along the lines of neural MCTS.http://proceedings.mlr.press/v80/guez18a.html …
- 2 more replies
New conversation -
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.