http://robowiki.net/w/index.php?action=history&feed=atom&title=Thread%3AUser_talk%3ABeaming%2FSmart_bots_competition%2Freply 2026-04-21T12:38:45Z Revision history for this page on the wiki MediaWiki 1.34.1 http://robowiki.net/w/index.php?title=Thread:User_talk:Beaming/Smart_bots_competition/reply&diff=33873&oldid=prev 2014-10-13T17:03:15Z

<p>Reply to <a href="/wiki/Thread:User_talk:Beaming/Smart_bots_competition" title="Thread:User talk:Beaming/Smart bots competition">Smart bots competition</a></p> <p><b>New page</b></p><div>I like the idea of more processing time in order to attempt more complex stats, but to be honest I've tried some pretty complex stuff (including [http://en.wikipedia.org/wiki/Spectral_clustering Spectral Clustering]) for KNN on recorded data and nothing has beat simple KNN with a square kernel and weighted on sample distance. Not to mention that a lot of these algorithms aren't just a constant 50x slower or whatever, they essentially become unsolvable at sizes above ~500 data points, scaling quadratically or worse.<br /> <br /> A problem with unlimited processing time is that it becomes possible to simply keep a copy of other bots and simulate them to determine where they will shoot/move. <br /> <br /> Does anybody have any specific techniques that they would use if there was more processing time available?</div>

Skilgannon