Score: 1

Tight Lower Bound for Multicolor Discrepancy

Published: April 25, 2025 | arXiv ID: 2504.18489v2

By: Pasin Manurangsi, Raghu Meka

BigTech Affiliations: Google

Potential Business Impact:

Makes fair sharing of items more difficult.

Business Areas:
A/B Testing Data and Analytics

We prove the following asymptotically tight lower bound for $k$-color discrepancy: For any $k \geq 2$, there exists a hypergraph with $n$ hyperedges such that its $k$-color discrepancy is at least $\Omega(\sqrt{n})$. This improves on the previously known lower bound of $\Omega(\sqrt{n/\log k})$ due to Caragiannis et al. (arXiv:2502.10516). As an application, we show that our result implies improved lower bounds for group fair division.

Country of Origin
🇺🇸 United States

Page Count
11 pages

Category
Computer Science:
Discrete Mathematics