Hi, I'm Liyana.

Researcher, Data Scientist · Mastercard AI Garage

I work on machine learning research at Mastercard's AI Garage, at the intersection of applied systems and publishable science. My current work focuses on representation learning, data-efficient training, and understanding what makes models memorize versus generalize, especially under noisy labels, class imbalance, and limited supervision.

I did my MTech in AI at IISc Bangalore, under Prof Soma Biswas. Before that, I was at GCE Kannur, where I first got curious about how machine learning models actually learn.

I also write: research notes, personal essays, and about the small details of ordinary life that somehow feel worth noting down.

Google Scholar GitHub LinkedIn Newsletter

Selected publications

CIKM 2025
Towards Equitable Coreset Selection: Addressing Challenges Under Class Imbalance
Liyana Sahir Kallooriyakath*, AN Reddy, BS Achary, A Sharma, K Shah, et al.
WACV 2024
AMEND: Adaptive Margin and Expanded Neighborhood for Efficient Generalized Category Discovery
A Banerjee, Liyana Sahir Kallooriyakath, S Biswas
ICIP 2024
AdaPrompt: Prompt Tuning with Adaptive Neighbours for Generalized Category Discovery
Liyana Sahir Kallooriyakath*, A Banerjee, S Biswas
ICDMW 2023
Study of Topology Bias in GNN-based Knowledge Graph Algorithms
A Surisetty, A Malhotra, D Chaurasiya et al.

All publications →

Recent writing

All writing →