Sourav Sahoo

I am a first-year PhD student at the Operations Research Center at MIT. I graduated with a dual degree (B.Tech + M.Tech) from the Department of Electrical Engineering at the Indian Institute of Technology Madras, Chennai where I was fortunate to be advised by Abhishek Sinha. My primary research interests are optimization, machine learning theory, and online learning.

EmailGoogle Scholar LinkedIn Twitter GitHub CV

News

[2023.09] Started PhD at MIT!

[2022.09] Our paper on online subset selection is out.

[2022.07] Graduated from IIT Madras!

[2022.02] Our paper on distributed online optimization was accepted at ACC 2022.

[2022.01] Our paper on k-experts problem was accepted at AISTATS 2022.

Preprints
Online Subset Selection using α-Core with no Augmented Regret
Sourav Sahoo, Siddhant Chaudhary, Samrat Mukhopadhyay, and Abhishek Sinha
Preprint

Conferences
Distributed Online Optimization with Byzantine Adversarial Agents
Sourav Sahoo, Anand Gokhale and Rachel Kalpana Kalaimani
American Control Conference (ACC), Atlanta, United States, June 2022.
Paper / BibTex

k-experts: Online Policies and Fundamental Limits
Samrat Mukhopadhyay, Sourav Sahoo, and Abhishek Sinha
International Conference on Artificial Intelligence and Statistics (AISTATS), Virtual, March 2022.
Paper / Code / BibTex

A Segment Level Approach to Speech Emotion Recognition using Transfer Learning
Sourav Sahoo, Puneet Kumar, Balasubramanian Raman and Partha Pratim Roy
Asian Conference on Pattern Recognition (ACPR), Auckland, New Zealand, November 2019.
Paper / Supplementary / Poster / Code / BibTex

Workshops
Multi-Modal Detection of Alzheimer's Disease from Speech and Text
Amish Mittal*, Sourav Sahoo*, Arnhav Datar*, Juned Kadiwala*, Hrithwik Shalu and Jimson Mathew
International Workshop on Data Mining in Bioinformatics (BIOKDD), Virtual, August 2021.
ArXiv / BibTex


*Authors contributed equally

Other Projects

SVRG-SO: SVRG for Stochastic Optimization
Sourav Sahoo
CS6515 Final Project
Report

Convergence and Implicit Regularization of Stochastic Mirror Descent in Overparameterized Models
Sourav Sahoo
EE5121 Term Paper
Report / Code