How to Add Pseudo Code to Moa

121 views
Skip to first unread message

Osama A.Mahdi

unread,
Jun 23, 2017, 3:02:31 AM6/23/17
to MOA users
Hi everyone, 

I am new to using Moa and I have wrote an algorithm, my question is how can I add or write the algorithm inside Moa to be implemented. 

I have read the Manual but I did not get it, its all about selecting classifiers and settings but did not show how to add algorithm.


Your help is highly appreciated. 

Yours truly,
Osama




Heitor Murilo Gomes

unread,
Aug 23, 2017, 5:48:55 AM8/23/17
to MOA users
Hi Osama,

To add a new algorithm what you need to do is extend from the appropriate Abstract class and implement the required methods. For example, if you are going to add a new classifier you will extend from AbstractClassifier and implement resetLearningImpl, trainOnInstance, ...
My advice is that you base your implementation on an existing classifier, for example, use NaiveBayes.java or OzaBag.java implementation as example.

Best Regards,
Heitor
Message has been deleted

Osama A.Mahdi

unread,
Aug 29, 2017, 2:23:58 AM8/29/17
to MOA users
Hi Heitor,

Many thanks for your reply and advice,

Do you mean implement my work using the existing classifiers with Java platform??? actually that what I have done. 

In this direction, plz do you have any idea how to train a classifier incrementally? I have found that Classifiers implementing the weka.classifiers.UpdateableClassifier to train incrementally. But who can I do it with MOA?

Your help is highly appreciated. 

Yours truly,
Osama


Heitor Murilo Gomes

unread,
Sep 3, 2017, 9:31:53 AM9/3/17
to MOA users
Hi Osama, 

You can do it in MOA by extending the AbstractClassifier class. Take a look at this tutorial at the MOA website:

This shall help you develop your own incremental classifier. 

Best Regards, 

Heitor

Osama A.Mahdi

unread,
Sep 13, 2017, 4:29:51 AM9/13/17
to MOA users
Dear Heitor,

Many thanks for your prompt reply, that's helped me indeed.


Regards 
Osama
Reply all
Reply to author
Forward
0 new messages