PhD Defended · March 2026 · IIT Gandhinagar

Anant Kumar

CS Researcher · Graph Theory, Algorithms & GNNs

I am a doctoral researcher at IIT Gandhinagar, supervised by Prof. Bireswar Das. My research spans graph isomorphism, subgraph counting, space-efficient algorithms, and the expressive power of Graph Neural Networks. I am currently seeking faculty positions in Computer Science & Engineering.

📄 Curriculum Vitae ✉ Email Me arXiv
Anant Kumar
🎓 Ph.D. CSE, IIT Gandhinagar
Thesis Defended · Mar 2026
🏛️ Asst. Prof., Mathematics
Saharsa College of Engineering
📍 Saharsa, Bihar, India
✉ kumar_anant@iitgn.ac.in
Latest
News & Updates
Mar 2026
🎉 Successfully defended my PhD thesis at IIT Gandhinagar. Degree awaited.
2025
📄 "Finding and Counting Patterns in Sparse Graphs" published in Journal of Computer and System Sciences (Elsevier).
Dec 2024
📄 "The Isomorphism Problem of Power Graphs and a Question of Cameron" at FSTTCS 2024, IIT Gandhinagar.
2024
📄 "Linear Space Data Structures for Finite Groups" published in Algorithmica (Springer).
2024
📄 "The Isomorphism Problem of Power Graphs and a Question of Cameron" accepted at FSTTCS 2024.
Work
Research
Graph Isomorphism Weisfeiler-Leman Algorithm Subgraph Counting Parameterized Algorithms Graph Neural Networks Space-Efficient Algorithms Algebraic Graph Theory Sparse Graph Algorithms Design & Analysis of Algorithms
My doctoral research centers on the design and complexity analysis of algorithms for fundamental graph-theoretic problems. I investigate graph isomorphism on algebraically structured graphs, develop space-efficient and parameterized algorithms for subgraph counting in sparse graph families, and explore the expressive power of Graph Neural Networks via structural localization. A recurring theme across my work is using algebraic and structural properties of graphs to circumvent worst-case intractability and design practically efficient algorithms.
Key Contributions
01
We developed polynomial-time algorithms for Graph Isomorphism on power graphs of finite groups, resolving a question posed by Cameron.
02
We introduced novel structural parameters — matched treewidth, matched treedepth, and DAG treedepth — and established their algorithmic utility for subgraph counting.
03
We designed constant-space and polynomial-space algorithms for counting patterns in degenerate and sparse graph families, improving best-known algorithmic exponents.
04
We Constructed linear-space, constant query-time data structures for finite groups with applications to algebraic computation.
05
We Proposed localization-based techniques to improve GNN expressivity for subgraph counting tasks.
Papers
Publications
Journal Articles
Linear Space Data Structures for Finite Groups with Constant Query-Time Journal
B. Das, A. Kumar, S. Sharma, D. Thakkar
Algorithmica, Springer · 2024
Finding and Counting Patterns in Sparse Graphs Journal
B. Komarath, A. Kumar, S. Mishra, A. Sethia
Journal of Computer and System Sciences (JCSS), Elsevier · 2025
Refereed Conference Papers
The Isomorphism Problem of Power Graphs and a Question of Cameron Conference
B. Das, J. Ghosh, A. Kumar
FSTTCS 2024 — 44th IARCS Annual Conference · LIPIcs, Schloss Dagstuhl
Linear Space Data Structures for Finite Groups with Constant Query-Time Conference
B. Das, A. Kumar, S. Sharma, D. Thakkar
STACS 2022 — 39th International Symposium on Theoretical Aspects of Computer Science
Finding and Counting Patterns in Sparse Graphs Conference
B. Komarath, A. Kumar, S. Mishra, A. Sethia
STACS 2023 — 40th International Symposium on Theoretical Aspects of Computer Science
Manuscripts Under Review
Counting Patterns in Degenerate Graphs in Constant Space Under Review
B. Komarath, A. Kumar, A. Pareek
Under review at a peer-reviewed venue
Improving Expressivity of Graph Neural Networks using Localization Preprint
A. Kumar, S. Das, S. Roy, B. Maity, A. Dasgupta
arXiv preprint
Workshop Papers
Local Fragments, Global Gains: Subgraph Counting using Graph Neural Networks Workshop
S. Roy, S. Das, B. Maity, A. Kumar, A. Dasgupta
Workshop on Differentiable Learning of Combinatorial Algorithms · NeurIPS Workshop
Education
Teaching
Assistant Professor, Mathematics
Saharsa College of Engineering, Bihar (AKTU-affiliated) · 2023 – Present
  • Engineering Mathematics I, II & III
  • Discrete Mathematics (B.Tech. students across departments)
  • Designed syllabi, question papers & outcome-based evaluation frameworks
  • Mentored students for university exams & undergraduate maths competitions
Teaching Assistant
IIT Gandhinagar · 2019 – 2023
  • Introduction to Computing ES102 (2019)
  • Discrete Mathematics CS322 (2020)
  • Algorithms CS610 (2020)
  • Data Structures & Algorithms II CS301 (2021, 2022, 2023)
  • Theory of Computation CS302 (2021, 2022)
Courses Prepared to Teach
Design & Analysis of Algorithms Data Structures Theory of Computation Discrete Mathematics Parameterized Algorithms Computational Complexity Theory Graph Theory Algebraic Methods in CS Engineering Mathematics Introduction to Machine Learning
Recognition
Awards & Achievements
🏆
CSIR-NET Junior Research Fellowship (JRF) — 2019
All India Rank 89 · Percentile 99+ · Mathematical Sciences
🎯
GATE 2019 — Mathematics
All India Rank 354 · Percentile 96+
🎖️
NCC 'A' Certificate — Grade A
Demonstrated leadership and discipline through National Cadet Corps training
Talks & Presentations
Invited Talk — Science Awareness Week
IIT Gandhinagar · 2022
Academic Service
→ Student Volunteer, ACM India Annual Event (2019)
→ Volunteer Organizer, IRISS 2019 — Indian Research in Information Security Symposium
→ Volunteer, RTA 2020 Workshop (Research Trends in Algorithms), IITGN
→ Volunteer, GIAN Programme (Global Initiative of Academic Networks), IITGN
Get in touch
Contact
✉
Email
kumar_anant@iitgn.ac.in
📞
Phone
+91 8700665806
🏛️
Current Position
Assistant Professor (Mathematics)
Saharsa College of Engineering, Bihar
🔬
Research Affiliation
PhD Scholar (Thesis Defended)
IIT Gandhinagar, Gujarat