Group Testing: An Important Tool for Data Science

Instructor: Ajit Rajwade
Class Timings: 6th July through to 10th July 2025
Class venue: DESE, Room 106, 2nd Floor
Office: KR-118, KReSIT Building (1st Floor Wing A)
Email: ajitvr AT cse DOT iitb DOT ac DOT in


Textbooks and Resources



Date

Content of the Lecture

Assignments/Readings/Notes

6/7 (Mon)
  • Course content overview
  • Definition of group testing, Motivating examples for group testing
  • Overview of applications
  • Broad Taxonomy
6/7 (Mon)
  • Adaptive group testing
  • Dorfman's method with analysis
  • Binary splitting and generalized binary splitting
7/7 (Tue)
  • Non-Adaptive group testing
  • Concept of k-separable, k'-separable, k-disjunct pooling matrices
  • Non-adaptive algorithms: smallest satisfying set (SSS) and Combinatorial Orthogonal Matching Pursuit (DD)
  • Non-adaptive algorithms: definite defectives (DD), Sequential COMP (SCOMP), Linear Programming Relaxation (LPR)
7/7 (Tue)
  • Analysis of COMP, DD, SCOMP
  • Proof that random Bernoulli matrices are disjunct matrices
  • Distance decoding algorithm
8/7 (Wed)
  • Noise models in group testing
  • Noisy linear programming relaxation
  • Noisy COMP with analysis
8/7 (Wed)
  • Noise definite defectives
  • Noisy distance decoding and (k,e)-disjunct matrices
9/7 (Thurs)
  • Real-valued group testing: Tapestry algorithm
  • Overview of RTPCR
9/7 (Thurs)
  • Near-neighbor search using binary splitting
10/7 (Fri)
  • Prevalence rate estimation using group testing
  • Multi-label classificatio using group testing
  • Image Moderation using group testing