MYVerified academic identity
Profiles open in a new tabResearch record
Google Scholar metrics checked 2 August 2026.
Research, made visible
Switch among three conceptual views inspired by Dr. Yousaf’s work: manifold geometry, privacy-aware distributed learning, and spatiotemporal tensor recovery.
A conceptual view of high-dimensional samples settling into a connected low-dimensional manifold while local neighbourhoods remain visible.
Research programme
A single concern connects the programme: how to make difficult data more useful without losing geometry, reliability, or trust.
Classical Isomap can become computationally expensive and sensitive to noise or unstable neighbourhood graphs.
Explore this programmeTraffic datasets lose observations across location and time, weakening downstream analysis and forecasting.
Explore this programmeDistributed and sensitive data systems need useful learning without weakening privacy, trust, or control of cryptographic keys.
Explore this programmeSelected publications
A compact tensor framework for recovering randomly and non-randomly missing traffic data while modeling multimode spatial and temporal structure.
Read via DOIA structured account of Isomap’s principal limitations, the methods designed to address them, and directions for future manifold-learning research.
Read via DOIAn incremental denoising method combining Hessian LLE and Gaussian-process kernels, recognized with the conference Best Paper Award.
Read via DOIOpen to conversation
For collaboration, research opportunities, invited talks, or student enquiries, contact Dr. Yousaf directly.
Start a conversation
