Dr. Sayan Ranu

Dr. Sayan Ranu

Professor at Department of Computer Science and Eng.

About

Sayan Ranu holds the joint positions of Nick McKeown Chair Professor at the Department of Computer Science and Eng. and the Yardi School of AI at IIT Delhi. His research interests span the broad area of machine learning and data mining for graphs. Sayan obtained his PhD from the Department of Computer Science, University of California, Santa Barbara (UCSB) in March 2012. Sayan has received several awards and recognitions including ACM India Early Career Researcher Award, Honorable Mention, 2024, Indian National Science Academy (INSA) Associate Fellowship 2023, Mrs. Veena Arora Early Career Faculty Research Award of IIT Delhi 2023, Associate of the Indian Academy of Sciences (IASc) 2020, Teaching Excellence Award 2025, and most reproducible paper award at SIGMOD 2018. Sayan regularly serves in the program committees and review panels of prestigious conferences and journals and has been awarded Outstanding Reviewer Awards at WSDM 2021, VLDB 2022, ICLR 2024 and KDD 2025. Sayan has been granted 5 US patents. Sayan’s dissertation on graph based techniques for querying and mining molecular databases served as the seed idea behind a life science based R&D start-up focusing on advanced technology for drug discovery.

Combinatorial optimization problems arise in domains such as viral marketing, computational sustainability, and healthcare. Their NP-hard nature makes exact algorithms impractical, leading to reliance on handcrafted heuristics or neural combinatorial solvers. Heuristics require significant expert effort and often remain sub-optimal, while neural solvers depend on NP-hard ground-truth data, generalize poorly out of distribution, and require GPU-based inference, raising scalability and sustainability concerns.
 
We reframe combinatorial optimization as a problem of autonomous algorithm discovery. Instead of predicting solutions, we design an AI agent that synthesizes executable heuristics directly. Leveraging large language models for code generation and evolutionary search for refinement, the agent proposes, evaluates, and iteratively improves candidate algorithms as explicit programs. Crucially, this process does not rely on optimal supervision, but improves solution quality through objective-driven feedback.
 
The result is transparent, human-readable code that runs efficiently on CPUs, eliminating neural inference costs. By discovering algorithmic structure rather than fitting parameters, the approach promotes interpretability, cross-domain robustness, and computational sustainability, offering a scalable pathway toward self-improving combinatorial optimization solvers.
Read More →
View All Speakers

We use cookies essential for this site to function well. Please click to help us improve its usefulness with additional cookies. Learn about our use of cookies in our Privacy Policy & Cookies Policy.

Show details