Menu
Scaling Discrete Integration via SAT and CSP
Assistant Professor Kuldeep S. Meel
National University of Singapore
Project Description
Decision making with uncertain data is increasingly becoming common in today’s world. Given complex dependencies between modules in a
system, and considering the inevitability of noise in observations, the design of symbolic reasoning techniques that can reason about probabilistic nature of systems have emerged as core challenge in the design of AI systems. The objective of this proposal is to develop symbolic reasoning techniques to aid artificial intelligence (AI) systems deal with uncertainty. To this end, this proposal focuses on development of fundamental algorithms for the problem of discrete integration, which is one of central components of symbolic reasoning in modern AI systems.
Research Technical Areas
Search and constraint satisfaction
Knowledge representation and reasoning
Reasoning under uncertainty
Benefits to the society
Discrete integration forms the backbone of probabilistic reasoning, which has wide spread usage in autonomous vehicle, healthcare technologies. CSP and optimization is also at the core of combinatorial and decision problems which occur quite widely throughout Computer Science and are highly applicable to many sub-problems in the smart-nation context. In a nutshell, this project introduces both academic values to put Singapore in the forefront of AI research and practical values for a safe and secure realization of Singapore smart-nation initiatives.
Project's Publications
-
Mate Soos, Kuldeep S. Meel:
BIRD: Engineering an Efficient CNF-XOR SAT Solver and Its Applications to Approximate Model Counting. AAAI 2019: 1592-1599 -
Sourav Chakraborty, Kuldeep S. Meel:
On Testing of Uniform Samplers. AAAI 2019: 7777-7784 -
Supratik Chakraborty, Kuldeep S. Meel, Moshe Y. Vardi:
On the Hardness of Probabilistic Inference Relaxations. AAAI 2019: 7785-7792 -
Rahul Gupta, Shubham Sharma, Subhajit Roy, Kuldeep S. Meel:
WAPS: Weighted and Projected Sampling. TACAS (1) 2019: 59-76 -
Yash Pote, Saurabh Joshi, Kuldeep S. Meel:
Phase Transition Behavior of Cardinality and XOR Constraints. IJCAI 2019: 1162-1168 -
Shubham Sharma, Subhajit Roy, Mate Soos, Kuldeep S. Meel:
GANAK: A Scalable Probabilistic Exact Model Counter. IJCAI 2019: 1169-1176 -
Alexis de Colnet, Kuldeep S. Meel:
Dual Hashing-Based Algorithms for Discrete Integration. CP 2019: 161-176
Team's Principal Investigator
Assistant Professor Kuldeep S. Meel
School of Computing
National University of Singapore
Principal Investigator’s Core Research Technical Areas
- Search and constraint satisfaction
- Verification for AI systems
- Interpretable machine learning
Introduction of the Principal Investigator
Kuldeep S. Meel is an Assistant Professor in the Computer Science Department of School of Computing at National University of Singapore, where he holds Sung Kah Kay Assistant Professorship. He graduated from the Indian Institute of Technology, Bombay, with a Bachelor of Technology (with Honors) in Computer Science, as well as a M.S. and PhD in Computer Science. The broader goal of his research is to advance artificial intelligence techniques, which utilize ubiquity of data and formal methods, to enable computing to deal with increasingly uncertain real-world environments.
Recent Notable Awards
- Sung Kah Kay Assistant Professorship, July 2018 — Present
- 2018 Ralph Budd Award for research in Engineering. This award, established in 1935, is given annually for the best doctoral thesis in the School of Engineering at Rice University.
- Honorable mention for 2018 ACP Doctoral Dissertation Award
Team
Co-Principal Investigator
Assoc. Prof. Roland Yap
National University of Singapore
Research Areas:
- Heuristic search and optimization
- Search and constraint satisfaction
- Program reasoning and verification