Research Portfolio

Publications

My research develops computational and statistical methods for optimal transport, Wasserstein geometry, dimensionality reduction, matrix recovery, federated learning, and scientific data analysis.

Conference Papers

Peer-reviewed and proceedings publications in signal processing, machine learning, and distributed learning.

Proceedings 2026

Recovering Wasserstein Distance Matrices from Few Measurements

Muhammad Sohel Rana, Abiy Tasissa, HanQin Cai, Yakov Gavriyelov, and Keaton Hamm

Proceedings of IEEE ICASSP 2026

PDF Code
Conference Submission 2026

Neighbor Embeddings Using Unbalanced Optimal Transport Metrics

Muhammad Sohel Rana and Keaton Hamm

IEEE ICASSP 2026 submission

Manuscript
Published 2025

Empowering Clients: Self-Adaptive Federated Learning for Data Quality Challenges

Zahidur Talukder, Muhammad Sohel Rana, Keaton Hamm, and Mohammad Islam

IEEE EDGE 2025

Paper

Under Review and Submitted

Current manuscripts on optimal-transport embeddings, scalable learning, and federated learning.

Under Review Current

On Wasserstein Distance Neighbor Embeddings

Muhammad Sohel Rana, Trinh, Bui, and Keaton Hamm

Manuscript under review

Preprint
Submitted Current

FedSRC: Federated Learning with Self-Regulating Clients

Co-authored manuscript including Muhammad Sohel Rana

Submitted manuscript

Manuscript
Manuscript Current

Neighbor Embeddings Using Unbalanced Optimal Transport Metrics

Muhammad Sohel Rana and Keaton Hamm

Research manuscript

Manuscript

Other Publications

Published Earlier Work

Nanofluid Natural Convection in Trapezoidal Annuli

Co-authored publication including Muhammad Sohel Rana

Applied mathematical and computational modeling

Publication
Replace each href="#" with the correct PDF, DOI, publisher, arXiv, or GitHub link.