Monday, November 04, 2013

CSjob: PhD studentship Collaborative Filtering; Indraprastha Institute of Information Technology, Dehli, India

Angshul Majumdar, a long time reader, just sent me the following

Hi Igor

Hope you are doing well. I have a position for Phd candidate in the field of collaborative filtering. I can't offer a stipend competitive enough for students vying for North American or European schools. But it would be good if I can get a good candidate from south east Asia or the middle east.

Would you mind putting this up in your blog.

I am attaching the blurb with this mail.


Angshul Majumdar
Assistant Professor
Indraprastha Institute of Information Technology
Thanks AngshulAngshul also tells me that  European or US based applicants are welcome to apply. Here is the ad:

Position: One PhD position is available at Indraprastha Institute of Information Technology, Delhi, India in the field of Collaborative Filtering a.k.a recommender systems.
Stipend: 500 USD equivalent in Indian currency.

Nature of the project: Collaborative filtering is used for predicting ratings / recommendations for online retail businesses such as Amazon, Netflix, Hulu etc. Till date most of the research in this area has been concentrated on palliating the basic issues like how to deal with sparsity (arising out of  missing ratings) and how to efficiently compute the ratings / predictions.  What is the next phase of research? How can we improve and customize ratings based on prior information like age, demography or gender? How can we incorporate group similarity information into the prediction algorithm? How can we efficiently update (without re-computing from scratch) the recommendation system when new users or new items are added?

Pre-requisites: The candidate should be have a Master’s degree (for highly motivated undergraduate degree holders are welcome as well) in Electrical (Electronics) engineering or Computer Science or Applied Mathematics. He /she should have working knowledge in linear algebra, probability and optimization. 

A background in machine learning is highly desired but not mandatory. Also it is good, if the candidate is versed in the literature of compressed sensing, low-rank matrix completion and matrix factorization. 

Assistant Professor, Indraprastha Institute of Information Technology, Delhi

No comments: