I am using genetic data to classify the genes. I want to know which
classifier gives the best classification accuracy. I was guided through
literature readings that AutoWekaClassifier, will do a comparative search
and outcomes the best classifier. When I applied it out, I got that
RandomForest gave 100% accuracy using 10 cross validation. Note that the
AutoWekaClassifier didn't suggest any searching algorithm for my data
1- When I applied RandomForest manually it gave me 86.3%. Why there is a gap
between the two results?
2- Also, I noticed that when applying other algorithms manually such as SMO,
I am getting better results (89.1%). Why AutoWekaClassifier didn't present
SMO for example as the best classifier?
Sent from: http://weka.8497.n7.nabble.com/