Naheed Anjum Arafat

prof_pic.jpg

Naheed is a post-doctoral fellow at Howard University DoD Center of Excellence in AI & ML . He has extensive experience as a Research Fellow at Rolls-Royce@NTU Corporate Lab, Nanyang Technological University (NTU), Singapore (2021-2024). At RR@NTU Corp Lab, he contributed to cutting-edge advancements in graph-based AI for computational physics simulation. He holds a Ph.D. in Computer Science from the National University of Singapore (2020), specializing in hypergraphs (higher-order graphs) and topological data analysis.

Naheed’s expertise encompasses graph representation learning, adversarially robust learning, and hypergraph data mining, with applications in privacy, cybersecurity, and computational physics. His contributions have been recognized through publications in premier venues such as ICML, ICLR, AAAI, VLDB, among others as well as two patents granted by the UK and US Intellectual Property Office.

Curriculum Vitae


Services:

  • PC Member: CODS-COMAD 2025, CODS-COMAD 2024, TKDE 2023, TKDE 2021, SKIMA 2014
  • Session Chair: VLDB 2023
  • Reviewer: ICML 2025, ICLR 2025, ICDE 2025, LoG 2024, NeurIPS 2024, CIKM 2024, JACT, DASFAA 2020, DAWAK 2020, ICDE 2018, VLDB 2017, DEXA 2017

news

Jan 22, 2025 Our paper on Logical Consistency of LLMs in Fact-Checking has been accepted at ICLR 25 (acceptance rate 32.08%). Paper Link
Dec 9, 2024 Our paper on Adversarial robustness of GNNs has been accepted at AAAI 25 (acceptance rate 23.4%). Paper Link Slides
Sep 24, 2024 New arXiv paper on Adversarial robustness of GNNs ( Paper Link )
Jul 24, 2024 Paper on measuring and reducing uncertainty of uncertain graphs has been accepted at IEEE DSAA 2024 (acceptance rate 26%) ( Paper ) ( Slides ) ( Code )
Jun 12, 2024 Paper on improving the fidelity of data-driven GNN models for fluid flow prediction selected for Spotlight at ICML 2024 (Only 3.5 % of the accepted papers)
May 2, 2024 Paper on improving the fidelity of data-driven GNN models for fluid flow prediction accepted at ICML 2024 (Acceptance rate 27.5 %)

latest posts

Jun 20, 2024 ICML24 poster

selected publications

  1. ICML
    Finite Volume Features, Global Geometry Representations, and Residual Training for Deep Learning-based CFD Simulation
    Loh Sher En Jessica*, Naheed Anjum Arafat*, Wei Xian Lim, and 2 more authors
    2024
    * = equal contribution
  2. pVLDB
    Neighborhood-Based Hypergraph Core Decomposition
    Naheed Anjum Arafat, Arijit Khan, Arpit Kumar Rai, and 1 more author
    Proc. VLDB Endow., Jul 2023
  3. DEXA
    Hypergraph drawing by force-directed placement
    Naheed Anjum Arafat, and Stéphane Bressan
    Database and Expert Systems Applications, Jul 2017
  4. DEXA
    Topological Data Analysis with ε-net Induced Lazy Witness Complex
    Naheed Anjum Arafat, Debabrota Basu, and Stéphane Bressan
    Database and Expert Systems Applications, Jul 2019
  5. DEXA
    Construction and Random Generation of Hypergraphs with Prescribed Degree and Dimension Sequences
    Naheed Anjum Arafat, Debabrota Basu, Laurent Decreusefond, and 1 more author
    Database and Expert Systems Applications, Jul 2020