Md Joshem Uddin
Topological & Geometric Machine Learning

Md Joshem Uddin

Ph.D. Candidate in Mathematics at UT Dallas working on topological machine learning, graph representation learning, temporal modeling, and AI for power grid resilience and cybersecurity.

Topological ML Graph Learning Temporal Graphs Power Grid AI Cyberattack Detection
📄 Download CV GitHub Google Scholar Email
11+
Published Papers
4+
Under Review
2026
PhD Expected
AAAI, ICLR
Top Venues

About

Research profile, interests, and current focus

I am a Ph.D. candidate in Mathematics at The University of Texas at Dallas. My research develops topology- and geometry-informed machine learning methods for graph-structured and temporal data, with applications in power grid resilience, anomaly detection, and scientific machine learning.

  • Focus areas: topological and geometric ML, graph representation learning, temporal and spatio-temporal modeling.
  • Applications: power systems resiliency, cyberattack detection, and graph-based predictive modeling.
  • Methods: graph neural networks, transformers, topological encodings, and contrastive learning.

Research Interests

Topological & Geometric ML Graph Representation Learning Statistical ML & Data Science Relational Deep Learning Temporal/Spatio-Temporal Modeling Cyberattack Detection Smart Grid Resiliency Applied & Computational Mathematics

Research & Academic Experience

Selected positions and research highlights

Graduate / Ph.D. Researcher — University of Texas at Dallas

2022 – Present | Advisor: Prof. Baris Coskunuzer

Designed and implemented novel ML/DL architectures (e.g., TopoFormer, SCNode, T3Former) for static and temporal graph learning; contributed to NSF-funded projects on grid resilience and anomaly detection.

Lecturer (Research Track) — University of Dhaka

2018 – 2021

Taught applied mathematics courses and conducted research in numerical modeling, including option pricing and atmospheric modeling.

Graduate Researcher — University of Dhaka

2017 – 2018

Conducted M.S.-level research on atmospheric phenomena using numerical simulations and theoretical modeling.

Selected Publications

Featured papers (expand this section with full BibTeX or links)

Conference Talks & Posters

Talks, invited talks, and poster presentations

T3former: Temporal Graph Classification with Topological Machine Learning

AAAI 2026 — Singapore Expo, Singapore (Jan 20–26, 2026)

WISE-GNN: Enhancing GNNs with Wise Embedding and Topological Encoding

SIAM Conference on Geometric and Topological Approaches in Data Science and ML (Invited Talk), Baylor University, Oct 2024

Topological Transformers for Graph Representation Learning

AMS Southeastern Sectional Meeting (Invited Talk), Tulane University, Oct 2025

MP-Grid: Detecting Power Grid Outages with Topological Machine Learning

AMS Southeastern Sectional Meeting (Invited Talk), Tulane University, Oct 2025

ATD/AMPS Joint PI Workshop Posters

George Mason University (2023) and Washington, DC (2024)

Teaching & Mentorship

Courses, TA experience, and student supervision

Teaching

  • Lecturer (Research Track), University of Dhaka — Advanced Linear Algebra, Mathematical Methods, Mathematical Modeling in Biology and Physiology, and more.
  • Teaching Assistant, UT Dallas — Calculus I/II, Differential Calculus, Linear Algebra, Calculus of Several Variables.
  • Lecturer, AIUB — Engineering Mathematics and ODE.

Mentorship

  • Supervised undergraduate researchers from Texas A&M and Virginia Tech on graph classification with TDA/ML (UT Dallas, Summer 2025).
  • Supervised undergraduate researcher from UT San Antonio on molecular property prediction using topological ML (UT Dallas, Summer 2024).
  • Guided multiple students on numerical modeling of stock prices (University of Dhaka, 2020).

Contact

Connect for research collaborations, talks, and academic opportunities