[mlpack] Alternatives to neighborhood-based collaborative filtering

[Ankush Mishra]

Hey mlpack,

I'm Ankush Mishra, currently a 4th year, in India, I was interested in 
implementing Alternatives to neighborhood-based collaborative filtering 
[1]. I have some limited experience with kNN's and Recommender Systems. 
Currently I'm working on Machine Learning and Computer Vision related 
problems. I was also a part of GSoC 2016, but under Apache Openmeetings[2]

I would have liked to implement the "Fast k-centers Algorithm"[3], but I 
see that Ankit has already got an implementation up. So, I'll be happy 
to be able to work on this.

Could someone help me start out contributing to this project? It's the 
first time for me, I have gone through the Mlpack, getting involved page.

Also, regarding this, is it just implementing the algorithm, based on 
the paper mentioned[4], or, as mentioned in [1], I will be implementing 
multiple of them, "weighted nearest neighbors and regression techniques" 
? Is there a list of these to be more specific, or should I build that 
up as well?

Sincerely

Ankush Mishra

[1] 
https://github.com/mlpack/mlpack/wiki/SummerOfCodeIdeas#alternatives-to-neighborhood-based-collaborative-filtering

[2] 
https://summerofcode.withgoogle.com/archive/2016/projects/5864271036547072/

[3] 
https://github.com/mlpack/mlpack/wiki/SummerOfCodeIdeas#fast-k-centers-algorithm--implementation

[4] 
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.129.4662&rep=rep1&type=pdf

-- 
Ankush Mishra