Publications

More Publications

. IMLI: An Incremental Framework for MaxSAT-Based Learning of Interpretable Classification Rules . In Proceedings of AAAI/ACM Conference on AI, Ethics, and Society(AIES)., 2019.

PDF

. BIRD: Engineering an Efficient CNF-XOR SAT Solver and its Applications to Approximate Model Counting. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI-19), 2019.

PDF Code

. BOSPHORUS: Bridging ANF and CNF Solvers . In Proceedings of Design, Automation, and Test in Europe(DATE),, 2019.

PDF

. On testing of Uniform Samplers . In Proceedings of AAAI Conference on Artificial Intelligence (AAAI-19)., 2019.

PDF

. On the Hardness of Probabilistic Inference Relaxations. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI-19), 2019.

PDF

. Knowledge Compilation meets Uniform Sampling. In Proceedings of International Conference on Logic for Programming Artificial Intelligence and Reasoning (LPAR)., 2018.

PDF Code

News

  • 6 December 2019

    Our paper on interpretable classification rules is accepted for publication at AIES 2019. Congratulations Bishwa on his first paper during PhD! Authors: Bishwamittra Ghosh and Kuldeep S. Meel

    30 November 2018

    The next version of ApproxMC is here: ApproxMC3 is faster and better than ever before. Do read Mate’s post for quick synopsys of the magic sauce behind ApproxMC’s performance.

    30 November 2018

    Congratulations Alexis for a successful defense of Masters thesis

    04 Novemeber 2018

    Kuldeep was awarded Distinguished Program Committee Member for IJCAI 2018.

    01 Novemeber 2018

    3 Papers accepted at AAAI-19!
    The first paper describes a new architecture for CNF-XOR formulas, which has allowed us to have a new approximate counter that is at least 100 times faster than ApproxMC2 – credit to Mate Soos for his breakthrough engineering that enabled this project!
    The second paper proposes the first algorithmic framework to test the distribution of a sampler.We show that several samplers output distributions far from what they claim!
    The third paper shows that several relaxations of probabilsitic inference do not give any computational advantage, raising concerns about the motivations behind their usage.

    24 September 2018

    Our paper on uniform sampling based on knowledge compilation is accepted at LPAR-22. Authors: Shubham Sharma, Rahul Gupta, Subhajit Roy, and Kuldeep S. Meel

People

Faculty

Kuldeep Kuldeep Meel

PhD Students

Yash Yash Pote
Mate Bishwamittra Ghosh

Masters Students

Shubham Shubham Sharma
Rahul Rahul Gupta

Research Assistant

Mahi Mohimenul Kabir

Undergraduates

Andre Andre
Suwei Suwei

Alumni

Mate Mate Soos
Bhavishya Bhavishya
Alexis Alexis de Colnet

Past visitors

Contact

Kuldeep S. Meel
Assistant Professor
Computer Science Department, School of Computing
National University of Singapore

Please consult Kuldeep’s calendar before suggesting a meeting time.

Openings

  • Looking for candidates with strong background in CS/Mathematics/Physics for post-doc for the project: Beyond NP Revolution. Read advertisement for more details.
  • Two post-doc positions available in the broad area of applying machine learning to SAT solvers, approximate counting techniques, and CP. Read advertisement for more details.
  • We are always looking for highly motivated Ph.D. students, research assistants with time commitment of at least 6 months and summer internship for exceptional undergraduate interns in our group.
    • If you are a student at NUS, feel free to drop by Kuldeep's office or schedule a meeting with him. (See his calendar).
    • Otherwise, please send him an email with your CV if you are interested. If you are looking for an internship/postdoc, make sure your subject contains the word "olleh" and you should include reviews of two of our group's papers (published in the previous 3 years). The reviews should be in the body of the email (and not as pdf). Furthermore, the body of your email should contain the phrase: "Here are two papers that I have reviewed". A strong background in statistics, algorithms/formal methods and prior experience in coding is crucial to make a significant contribution to our research.