Scholar
Shubhendu Trivedi
Google Scholar ID: EbyGwncAAAAJ
Massachusetts Institute of Technology
Citations & Impact
Citations
1,984
H-index
15
i10-index
20
Publications
20
Co-authors
20
list available
Contact
Publications
Resume (English only)
Education
- PhD completed in August 2018 with a thesis on similarity learning, metric estimation, and group covariant neural networks.
- Primary PhD advisor: Prof. Gregory Shakhnarovich at Toyota Technological Institute at Chicago (TTIC).
- Collaborated closely with Prof. Risi Kondor (University of Chicago) and Prof. Brian D. Nord (Kavli Institute/Fermilab, Deep Skies Lab) during PhD.
- Earned an MS focusing on Machine Learning prior to PhD candidacy.
- Earlier MS in Computer Science from Worcester Polytechnic Institute under Profs. Neil T. Heffernan and Gábor N. Sárközy, with thesis reader Sonia Chernova; thesis introduced a new clustering algorithm based on Szemerédi Regularity Lemma and a clustering-based ensemble method akin to mixture of experts.
Background
- Broad interests in Machine Learning, with a focus on (deep and otherwise) representation learning, structured prediction, and semi/weakly/self-supervised learning.
- Currently exploring supervised similarity and distance learning in low-shot regimes and representation learning for combinatorial structures like graphs and sets.
- Designing neural architectures with task-relevant symmetries using group and representation theory (group-equivariant neural networks) for principled and data-efficient design.
- Draws inspiration from applications in computer vision and physical sciences, especially computational chemistry and physics.
- Maintains an amateur interest in extremal combinatorics and spectral graph theory.
