Modified EFDT

[darshna-creator]

[darshna-creator]

air i want to change to change base learner is it poasible?? if yes then how??

On Fri, 6 Mar, 2020, 2:01 am Heitor Murilo Gomes, [email protected] wrote:

@hmgomes commented on this pull request.

In moa/src/main/java/moa/classifiers/meta/StreamingRandomPatches.java https://github.com/Waikato/moa/pull/199#discussion_r388547301:

@@ -76,7 +76,7 @@ private static final long serialVersionUID = 1L;

 public ClassOption baseLearnerOption = new ClassOption("baseLearner", 'l',

•        "Classifier to train on instances.", Classifier.class, "trees.HoeffdingTree -g 50 -c 0.01");
  

•        "Classifier to train on instances.", Classifier.class, "trees.ARFFIMTDD -g 50 -c 0.01");
  

The default base learner for StreamingRandomPatches should not be ARFFIMTDD.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Waikato/moa/pull/199?email_source=notifications&email_token=AOTTEY6XWTZQZXZBKK52XJDRGADZDA5CNFSM4LCG3W4KYY3PNVWWK3TUL52HS4DFWFIHK3DMKJSXC5LFON2FEZLWNFSXPKTDN5WW2ZLOORPWSZGOCYGBUEI#pullrequestreview-369891857, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOTTEYYFH3YWE6B43OYVU5DRGADZDANCNFSM4LCG3W4A .

[hmgomes]

air i want to change to change base learner is it poasible?? if yes then how?? …

Yes, it is possible. You can do that in the GUI or in the command line, just change the parameter. For example, in the command line:

EvaluateInterleavedTestThenTrain -l (meta.StreamingRandomPatches -l bayes.NaiveBayes)

[darshna-creator]

sir how to use EFDT.java algorithm on stream.

[darshna-creator]

I want to efdt algorithm on gui ......but I am not able to find it

[darshna-creator]

got it

[darshna-creator]

I want to add MEFDT which is a modified EFDT in classifiers. Sir Kindly confirm about changes or let me how to test my algorithm on experimental tab of MOA. Please Anyone reply as soon as possible.