[mlpack] Regarding GSoC 2018 - VAE project

[Ryan Curtin]

On Fri, Feb 16, 2018 at 06:33:23PM +0530, Namrata Mukhija wrote:
> Hi
> 
> I'm Namrata, final year IT undergrad from NSIT, India. I'm new to the
> community, and would love to contribute to mlpack and participate in GSoC
> 2018.
> 
> I have gone through the ideas list for GSoC 2018 and implementing VAEs is
> the project I'm really interested in followed by Essential Deep Learning
> modules project. How should I go about it?
> 
> I have gone through get involved guide would love to get to know some
> pointers on the steps for getting started with the VAE project
> particularly.

Hi Namrata,

Welcome to the community!  The VAE project will be mentored by Sumedh,
so he probably has more input on the specific project, but I would say a
great way to get started would be to get familiar with the deep learning
code in mlpack.

Eugene Freyman wrote a really nice implementation of a network that
learns the MNIST digits for a Kaggle competition in the models
repository:

https://github.com/mlpack/models/

You could play with that, as well as the tests in src/mlpack/tests/, to
learn how the framework works and get some experience with it.  That
should help guide the design of VAEs.

I hope this helps---let me know if I can clarify anything I've written.

Thanks!

Ryan

-- 
Ryan Curtin    | "I like this game."
ryan at ratml.org |   - Coach