Deep learning for Group Testing

Basic Information

Project Title / Domain: Deep learning for Group Testing
Name: Shreyas Grampurohit
Guide: Satish Mulleti, Ajit Rajwade
Project Type: Both implementation as well as theory

Short Description of Project

Exploring interpretable/verifiable deep learning methods with generalisation guarantees for the group testing/compressed sensing problem

Whom did you work with?

The prof directly

Tools / Simulation / Software / Hardware

Python, Server access with GPUs

Expectations from Guide

Be updated with my progress, suggest technical solutions wherever possible/necessary; communicate/connect with experts in the field if that helps speed up the progress.

Expectations for 8th Sem & Summer

I worked in my 8th semester, so the guides were okay with me going for an internship

Load: 8th Sem vs 9th (Placement) Sem

I had a lot of DDP workload in my 9th sem, despite the placement semester, because I was aiming to submit multiple papers. Otherwise, there wasn’t too much pressure from the guides because I had worked in the summer as well.

Summer on Campus

It wasn’t the most productive for me, but it surely helped me progress enough to be able to aim for multiple paper submissions by September, although key progress happened during the semester.

Is DDP Guide Same as SRE/RnD Guide?

Yes

How did your SRE help with your DDP?

Mainly it helped with building the rapport with the guides. My topic effectively changed for my DDP.

End Deliverables

A working (interpretable) non-iterative (deep-learning) solution for the group testing problem, along with theoretical guarantees/generalisation bounds.

Advice

You’ll be meeting weekly. Communicate logistics well in advance; Always try to make every meeting productive, and keep in mind how it builds upon your previous work or whether you need to switch gears/route differently.