I am working on a classification problem, where I have 400 features(all are numeric), and 100 classes and I have 26,000 examples for training. In my project I am using Weka and I have tried different algorithms and so far I have received the best results with the following stack of algorithms:

  • Rotation Forest(with 60 Iteration)+Multiclass classifier(1-vs-all)+RepTree-> accuracy 79.9%
  • Bagging(with 60 Iteration)+Multiclass classifier(1-vs-all)+RepTree->accuracy is 79%

I have tried also to increase the number of Iterations but the accuracy decreases after 60 Iterations

Q: Is it possible to increase the accuracy, and which algorithm or methodology can be used to achieve better accuracy?

  • $\begingroup$ Have you tried feature selection? Weka offers a variety of tools to assist in feature selection. With 400 features, that might help. $\endgroup$ – G5W Mar 22 '17 at 13:51
  • $\begingroup$ Thanks for your reply. I have used the LMT classifier in weka and it gave me overall success rate around 93% $\endgroup$ – x_zero Mar 29 '17 at 9:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.