Sayan Bandyapadhyay
Portland State University, OR·Computer Science
Clustering means sorting data points into groups of similar items. It is one of the most common ways to analyze data, and its results often feed into machine learning systems.
- Fields
Computer Science
Mathematics & Statistics
- Topics
clustering
algorithmic fairness
k-means
- Grants
- 1 known grant, through Dec 2026
What they work on
- Design polynomial-time approximation algorithms for balanced k-median and k-means clustering, where each cluster must fairly represent every protected group, aiming for constant-factor guarantees like those known for ordinary clustering.
- Determine the limits and applicability of generic approximation techniques for balanced clustering, including relax-and-round, which solves a relaxed version of the problem and rounds the answer, and metric partitioning, which divides the point space into pieces.
- Develop new algorithmic tools for fair clustering that also extend to a broader class of related optimization problems.
- Characterize trade-offs among time complexity, approximation guarantee, and the degree to which fairness constraints are satisfied.
More about the research
When the data describe people, a plain clustering can leave some groups badly underrepresented, which can make downstream systems unfair to people defined by a sensitive feature such as gender. Fair clustering adds rules that prevent this. Designing fair clustering algorithms with mathematically proven performance has turned out to be much harder than for ordinary clustering, because the classic techniques fall short on the fair versions.
Why it matters
Clustering organizes data and shapes the machine learning systems built on it. Fair clustering algorithms with proven guarantees could help such systems treat each protected group, for example groups defined by gender, in an unbiased way. Mapping the trade-offs between speed, solution quality and fairness could help practitioners pick algorithms suited to how much fairness an application requires.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Viswanath Nagarajan
University of Michigan, MI·Computer Science
Design approximation algorithms with provable performance guarantees for sequential testing of systems made of interconnected components, including series-parallel structures, under complex cost structures.
Kirk Pruhs
University of Pittsburgh, PA·Computer Science
Develop combinatorial surrogates, graph-theoretic and local expansion conditions, that imply the high topological connectivity needed by topological fixed-point proofs but can be verified by algorithm researchers without topology training.
Nikhil Bansal
University of Michigan, MI·Computer Science
Determine how many iterations the interior-point method, a standard algorithm for large optimization problems, needs in the worst case, on average and on sparse inputs, to understand its true complexity.
The school
Portland State University: 3 labs on PI Pages, $52M a year on research, 92 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.