Aditya,
Please take a look at the plan in mahout-1929 specifically the list of next steps I identified to see which of those interests you and which you think you can finish in a managable period of time, for all GLM discussions lets keep all our discussions inside that JIRA. I am also in the middle of researching ALS so feel free to take a look at the discussion with Dmitry to see whether you want to work on any of the areas identified in that thread (namely (1) double-weighed ALS
(2) bayesian optimization and crossvalidation in an n-dimensional
hyperparameter space
(3) feature extraction per (preferrably R-like) formula.) , if you have questions ping the list and I'll respond as my bandwidth frees up some more.
Thanks again for your help.
________________________________
From: Aditya <***@gmail.com>
Sent: Tuesday, March 7, 2017 7:29 AM
To: ***@mahout.apache.org
Subject: Re: Regarding Issue MAHOUT-1941
Thanks for the advice Saikat. I'll try to get some hands on experience. I
have gone through the mail thread " Contributing an algorithm for Samsara "
and was wondering if there are any unit tests / any bug fixes pending to
which I could contribute.
I will also go through the list of issues in Jira and see if I can
contribute to any of them to understand the framework. I'll get back to you
if I have any questions.
Best Regards,
Aditya
Post by Saikat KanjilalAditya,
Welcome to mahout, you can start by taking a look at the plan I've
outlined in the JIRA to Trevor and others and add your comments there for
tracking purposes, I am looking to get a shell API up for review in its
very crude state hopefully in the next week or two based on the outlined
plan. Also depending on your experience level it may be worthwhile to
solve a few simple issues first which require getting your hands into the
code, for example unit tests are a good place to do that to understand the
Samsara framework before implementing any large algorithm. My 2 cents for
what its worth. Use the dev list for help as needed.
Thanks
________________________________
Sent: Tuesday, February 28, 2017 12:03 AM
Subject: Regarding Issue MAHOUT-1941
Hello Saikat,
I have been talking to Trevor in relation to contributing to Apache Mahout
in terms of Algorithm development / improvements and he suggested that
currently you and Jim are working on Generalized Linear Model.
I was looking through the list of issues and found issue 1941 and was
wondering if there is an opportunity to contribute there. In general, I
would be grateful if you could guide me on how I can make myself useful to
Apache Mahout.
Awaiting your reply.
Regards,
Aditya