Mathematician & Researcher
Pearl
Kuuridong
I work at the intersection of combinatorial optimization and graph-based learning, with a focus on how discrete structure shapes what algorithms can find.
Recently defended my MSc thesis at AIMS Rwanda on combinatorial optimization with graph neural networks. Previously a data engineer; actively pursuing doctoral study.
Education
MSc in Mathematical Sciences —
Data Science
AIMS Rwanda · 2025 – 2026
BSc in Mathematical Sciences — Mathematics & Computer Science
University of Ghana · 2018 – 2022Combinatorial Optimization with Graph Neural Networks
Applies the GCON self-supervised GNN framework (Wenkel et al., 2024) to two new combinatorial optimization problems: Maximum Independent Set and Balanced Graph Partitioning, by deriving problem-specific loss functions and rule-based decoders. A key finding is the tension between objective optimization and constraint satisfaction: spectral methods minimize cut size effectively but often violate balance constraints, while learning-based approaches better preserve feasibility.
Read Full Thesis (PDF)Applying Persistent Homology to Topological Data Analysis (TDA)
Undergraduate thesis introducing persistent homology as a tool for extracting topological features from data covering the foundations in point-set and algebraic topology, the persistent homology pipeline, and a worked Python example computing persistence diagrams for a point cloud using Rips complexes.
Read Full Thesis (PDF)-
Reading Jul 2026
-
Finished Apr 2026
-
Jul 25
Two Piles, One Golden Ratio
What backward induction reveals about Wythoff's Nim.
-
Jun 16
A Practical Approach to Denoising Sound
Using Fourier transforms to separate signal from noise.
Interested in research collaborations, doctoral opportunities, or just want to talk graphs and combinatorial optimization? I would love to hear from you.
Get in touch