Neural Sequence Generation with Constraints via Beam Search with Cuts: A Case Study on VRP
Pouya Shati, Eldan Cohen, Sheila McIlraith
About
I am a data scientist interested in interpretable and constrained machine learning through combinatorial optimization.
I got my PhD from the Computer Science department at the University of Toronto and Vector Institute. My Supervisors were Sheila McIlraith and Eldan Cohen. I did my undergraduate studies in Software Engineering at Sharif University of Technology.
Research Interests
Links
Research Summary
I am interested in utilizing combinatorial optimization, symbolic reasoning, and logical formalisms in machine learning. I have shown through my work that doing so enables interpretable machine learning, solution constraints, and ML models that are enhanced in rigorous reasoning capabilities. My work facilitates ML application to sensitive tasks where the model should be thoroughly understood and analyzed. It further supports tasks where domain-specific knowledge should be integrated in the solution. Lastly, it elevates common ML models such as LLMs by combining them with symbolic reasoning towards solving problems in domains such as planning, program synthesis, and vehicle routing.
Publications
SAT-Based Learning of Compact Binary Decision Diagrams for Classification
Pouya Shati, Eldan Cohen, Sheila McIlraith
Optimal Decision Trees For Interpretable Clustering with Constraints
Pouya Shati, Eldan Cohen, Sheila McIlraith
Slides, short slides, and poster
SAT-based optimal classification trees for non-binary data
Pouya Shati, Eldan Cohen, Sheila McIlraith
SAT-Based Approach for Learning Optimal Decision Trees with Non-Binary Features
Pouya Shati, Eldan Cohen, Sheila McIlraith
The Evolution of Indirect Reciprocity Under Action and Assessment Generosity
Laura Schmid, Pouya Shati, Christian Hilbe, Krishnendu Chatterjee
Get In Touch
I would love to hear from you on interesting points of discussion or potential collaborations!