http://robowiki.net/w/index.php?title=Thread:Talk:Pris/Dodging_Performance_Anomaly%3F/reply_(7)&feed=atom&action=historyThread:Talk:Pris/Dodging Performance Anomaly?/reply (7) - Revision history2024-03-29T15:03:44ZRevision history for this page on the wikiMediaWiki 1.34.1http://robowiki.net/w/index.php?title=Thread:Talk:Pris/Dodging_Performance_Anomaly%3F/reply_(7)&diff=32223&oldid=prevSkilgannon: Reply to Dodging Performance Anomaly?2013-11-16T21:27:54Z<p>Reply to <a href="/wiki/Thread:Talk:Pris/Dodging_Performance_Anomaly%3F/reply_(6)" title="Thread:Talk:Pris/Dodging Performance Anomaly?/reply (6)">Dodging Performance Anomaly?</a></p>
<p><b>New page</b></p><div>Recurrent NN uses a lot of memory and processing power, both of which are fairly limited in the RoboCode setting. Speed is definitely the main issue, particularly when a lot of the time is already taken doing predictions to give more relevant features for classification. Even with spreading calculations over multiple ticks many popular techniques Just Wouldn't Work.<br />
<br />
If you can't get multiple outputs out of the RF, just run a bunch of them, one in each bin, and choose the bin with the highest probability. Ie, each bin is a different class and you choose the most probable class. Quick and dirty regression without inter-dependency. I've actually thought about trying a Naive Bayes like this, just for kicks. I think Pris and a few others do their NN classifications this way.</div>Skilgannon