This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2016-11-21
Channels
- # beginners (165)
- # boot (9)
- # cider (9)
- # cljs-dev (5)
- # cljsjs (5)
- # clojars (4)
- # clojure (207)
- # clojure-brasil (1)
- # clojure-greece (3)
- # clojure-poland (2)
- # clojure-russia (6)
- # clojure-spec (85)
- # clojure-taiwan (1)
- # clojure-uk (53)
- # clojurescript (96)
- # community-development (2)
- # cursive (4)
- # datomic (14)
- # emacs (41)
- # events (2)
- # hoplon (184)
- # leiningen (1)
- # nginx (1)
- # off-topic (16)
- # om (7)
- # onyx (63)
- # pedestal (27)
- # planck (17)
- # protorepl (3)
- # rdf (9)
- # re-frame (62)
- # reagent (7)
- # ring-swagger (5)
- # schema (2)
- # spacemacs (5)
- # test-check (25)
- # untangled (93)
- # yada (3)
This is little Filip, one of my good friends from University. Five months ago he was a waiter in Philadelphia, PA. He saved up a couple thousand, packed everything up, and moved to Thailand. Little Filip started filming his big adventures, and the videos are pretty awesome. https://www.youtube.com/watch?v=CDTcSfw2MAc
@sveri, @fellshard, you might also want to look into reinforcement learning: https://deepmind.com/blog/deep-reinforcement-learning/ Though I guess you'd have to study a bit more for that than for entry-level genetic programming techniques. If you're interested, I recommend https://youtu.be/2pWv7GOvuf0.
> Maybe there was a bit too much content though..? I'm not sure. At places it felt like you're in a hurry. This is true. I did cram in a bit more than I should’ve.
is there anything we can do about the #admin-announcements channel? Every time I switch back to clojurians slack after a while of using something else, it focuses the #admin-announcements channel
@rmoehn Thanks for the pointer, I actually watched all the games Alpha Go played against Lee. The thing is in halite you only have on second per round. So I think I need a network that just produces instant output and not like what Alpha Go did. So I need a trained network and I am not sure if that is possible with RL? I think I need to grab more theory 🙂
So you give the algorithm some limited time to work, pop out the best answer it can come up with in that time. You can think of it as navigating a search space and trying to pick how to use that time to search wisely each round.
@sveri heavily recommend this lecture 🙂 https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-034-artificial-intelligence-fall-2010/lecture-videos/lecture-6-search-games-minimax-and-alpha-beta/
@val_waeselynck thank you 🙂
you're welcome. Actually, I recommend watching the whole course if you have the time, it's worth every minute.
(And I don't even work in AI)