Curriculum Vitae

Dallas, TX · sxk230046@utdallas.edu

Download CV (PDF)

Professional Summary

Ph.D. Candidate in Mathematics specializing in Topological Data Analysis and Graph Representation Learning for biomedical applications, including brain network modeling and ligand-based virtual screening (LBVS) for drug discovery. Experienced in CUDA-enabled GPU programming and scalable model development on high-performance computing (HPC) clusters.

Education

• Ph.D. in Mathematics, The University of Texas at Dallas (Expected May 2028) · GPA: 3.75/4
• M.S. in Mathematics (Data Science Specialization), The University of Texas at Dallas (Expected Dec 2026) · GPA: 3.75/4
• Master of Mathematics, National Institute of Technology, Rourkela (May 2017) · GPA: 7.87/10
• B.S., Mathematics, Computer Science, Physics, Guru Nanak Dev University (June 2015) · GPA: 7.9/10

Technical Skills

• Programming: Python, C, C++, Java, SQL, R, MATLAB, Mathematica
• Libraries/Frameworks: PyTorch, PyTorch Geometric, DGL, TensorFlow, scikit-learn, NetworkX, RDKit, XGBoost, NumPy, SciPy, Gudhi, Ripser
• Tools/Platforms: CUDA, Slurm (HPC), Linux, Neo4j, Git, Databricks, Jupyter Notebook, LaTeX, MS Office

Projects

• TopU-LBVS: A Realistic Multi-Target Benchmark for Ligand-Based Virtual Screening (In Progress)
  Constructed the unbiased TopU95 multi-target dataset from ChEMBL35 and established a rigorous benchmarking protocol. Evaluated naive models, classical ML baselines, deep learning graph models, and foundation models under a unified framework.
• Brain Network Graph Classification (Under Review, ICML 2026)
  Analyzed how atlas choice affects downstream performance and introduced edge-based quadratics to improve structural expressiveness, validated through multi-atlas baselines with consistent gains.
• Low-Shot Graph Learning with Topological and Spectral Embeddings (Accepted, LoG 2025)
  Introduced a prototype-based low-shot graph learning model using Betti vectors and spectral DoS embeddings, outperforming GNN and Graph Transformer models.

Experience

• Teaching Assistant, The University of Texas at Dallas (Aug 2023 – Present)
  Conduct problem-solving sessions and oversee grading for calculus courses; served as Head TA for Calculus of Several Variables.
• Subject Matter Expert – Advanced Mathematics, Ed-Tech Industry (Remote) (Jun 2017 – Jul 2023)
  Developed advanced mathematics content for platforms including Chegg and Meritnation.

Publication

• Kumar, S., Saha Ray, S. Numerical treatment for Burgers–Fisher and generalized Burgers–Fisher equations. Math Sci 15, 21–28 (2021).