Aravind Reddy
Aravind Reddy
Home
News
Research
Publications
Funding
Teaching
Mentoring
CV
Contact
Light
Dark
Automatic
3
Beyond the First Hit: Towards Tight Guarantees for k-means++
k-means++ is the standard way to start k-means clustering. Its worst-case approximation ratio was known to be at most 8(ln k + 2), improved to 5(ln k + 2) in our NeurIPS 2020 paper.
Aravind Reddy
PDF
Cite
arXiv
Machine Unlearning for Federated Socially Fair k-Means Clustering
Hospitals that cluster patients jointly, without pooling records, must treat demographic groups fairly and honour deletion requests. We build k-means++ summaries for each institution and group, prove an O(log² k) guarantee that does not depend on how groups are split across institutions, and show that replay reproduces retraining exactly and 35.
Aravind Reddy
,
Deeraj S K
,
Mohd Wasif Raza
,
Saurav Prakash
PDF
Cite
Code
arXiv
Pluralis v0.1: Towards a Multicultural, Multimodal, Multilingual Benchmark for AI Risk and Reliability
Pluralis, part of the MLCommons AILuminate effort, is a culture-first benchmark of 6,448 image-text prompts across six Asia-Pacific locales and eight languages. Each prompt pairs innocuous text with a benign image that together trigger a local legal or cultural violation, exposing failures that globally averaged metrics hide.
Alicia Parrish
,
Rajat Shinde
,
and others
,
Aravind Reddy
,
Sunayana Sitaram
,
Balaraman Ravindran
,
Lora Aroyo
PDF
Cite
arXiv
Cite
×