Orange Forum • View topic - GSOC 2011

GSOC 2011

General discussions about Orange and with Orange connected things (data mining, machine learning, bioinformatics...).

GSOC 2011

Postby Somya » Sun Mar 20, 2011 23:15

Hello ,

I'm a 4rth year( B tech + MS) student in Computer Science at IIIT Hyderabad,India. I have a very good background of programming languages like C, C++ and Python.

I'm interested in working on some of the ideas (Add built-in attribute weighting support and Improving K-nearest neighbors) in the list .I was just wondering how and whom to submit the idea proposal. Is there any template for idea submission? Are there some mentors assigned to these projects? I couldn't find the information related to same on the site . Kindly Guide me.

Thank You,
Somya Srivastava

Re: GSOC 2011

Postby Mitar » Mon Mar 21, 2011 2:11

The idea proposals should be submitted through Google web application (once the submission opens on March 28). There is also a template provided for that. But you can find template also on this page.

If you want to discuss your proposal in advance please use this forum or join us on our Skype chat.

Re: GSOC 2011

Postby Mitar » Thu Mar 24, 2011 9:28

Just to warn you that we have removed built-in attribute weighting support and Improving K-nearest neighbors from our ideas list few days ago after we discovered that we could not mentor this properly.

Re: GSOC 2011

Postby hamzah » Fri Mar 25, 2011 15:12

Hi everybody,
First let me tell you about my experience.
I am a master student (expected to graduate this May) in Computer Science major.
I am working in Brain Computer Interface (BCI) classification and feature extraction algorithms.
I used Orange for Data mining course project last year. and this semester I am working in python for my AI course projects.
I do have good experience in OOP (C++,Java, Python and Matlab).
Second I read through the ideas suggested by you and I liked the Multi-Class classification
Third I wish to have a toolbox for some algorithms for BCI which can be used for other research topics also.
some of these algorithms: Common Special filter (CSP), FIR and IIR filters, Log Variance feature extraction and other algorithms.
Such algorithms will be useful for researchers in machine learning, specially Orange is a friendly toolbox.

Another suggestion is to make a port connection between Orange and other known software like Matlab to make it easier for people to transfer or test their work in orange.

I wish to have further discussion about these suggestions. And to know more about student selection process.

Thanks :)

Return to General Discussions