weka - what is the cause of Accuracy difference? -


i have dataset of 2001 attribute , 63 instances while doing classification using weka can see j48 have higher accuracy compared naive bayes.but after doing pca can see j48 accuracy decrease compared first 1 first got 82% , after pca got 72.5%.what reason?

decision trees select features.

pca mixes features components.

this makes feature selection work less - every input feature influences (almost) every derived feature.

don't assume pca magic tool solves problems. can work - can make things worse.


Comments

Popular posts from this blog

node.js - How to mock a third-party api calls in the backend -

node.js - Why do I get "SOCKS connection failed. Connection not allowed by ruleset" for some .onion sites? -

matlab - 0-by-1 sym - What do I need to change in order to get proper symbolic results? -