search

Waikato / meka

Multi-label classifiers and evaluation procedures using the Weka machine learning framework.
http://waikato.github.io/meka/
GNU General Public License v3.0
200 stars 76 forks source link

MULAN Wrapper #61

Open FernandoSaez95 opened 5 years ago

FernandoSaez95 commented 5 years ago

Hello,

Can I add more MULAN methods to the wrapper?

I have seen there is only possible to use BR, LP, CLR, RAkELn, MLkNN, IBLR_ML, BPMLL and HOMER. What should I do if I would want to use for example Ensemble Classifier Chains of MULAN?

Thank you.

jmread commented 5 years ago

Hi,

Ensemble of Classifier Chains is already implemented in MEKA. If you want to use the MULAN version you will need to write a new entry for MULAN.java, e.g., else if (m_MethodString.equals("ECC")) { ... etc.

Also note that BR, LP, RAkEL,. are also implemented directly in MEKA.

FernandoSaez95 commented 5 years ago

Thank you!!! So helpful.

Could you please tell me which method of MEKA API should I use for Ensemble of Classifier Chains?

jmread commented 5 years ago

One example would be meka.classifiers.multilabel.meta.BaggingML using meka.classifiers.multilabel.CC.

m-bittencourt commented 5 years ago

Hello,

I need to work with MMPLearner in incremental learning. Is it possible to work with Mulan wrapper in this scenario?

Thank you.