Hi,
I tried to find SNAP codes for classification part but I did not find them. Would you please guide me?
Thanks @lveci.
1.I need to know what algorithm exactly is behind the RF ranking in SNAP but I could not find it in the source code.
Something like this;
http://blog.datadive.net/selecting-good-features-part-iii-random-forests/
Do you know where is that or there is any documentation about the ranking in SNAP?
@ABraun had a discussion here;
I do not know, if he can help us or not?
-
RF also provide two straightforward methods for feature selection: 1. mean decrease impurity (MDI) 2.mean decrease accuracy(MDA); which one was used in SNAP?
-
I have some values in my result but I do not know what is the meaning with them like 'score: tp (true positive),accuracy, precision, correlation
, errorRate,cost, GainRatio .