English

Complexity-Theoretic Implications of Multicalibration

Computational Complexity 2024-07-30 v2 Computers and Society

Abstract

We present connections between the recent literature on multigroup fairness for prediction algorithms and classical results in computational complexity. Multiaccurate predictors are correct in expectation on each member of an arbitrary collection of pre-specified sets. Multicalibrated predictors satisfy a stronger condition: they are calibrated on each set in the collection. Multiaccuracy is equivalent to a regularity notion for functions defined by Trevisan, Tulsiani, and Vadhan (2009). They showed that, given a class FF of (possibly simple) functions, an arbitrarily complex function gg can be approximated by a low-complexity function hh that makes a small number of oracle calls to members of FF, where the notion of approximation requires that hh cannot be distinguished from gg by members of FF. This complexity-theoretic Regularity Lemma is known to have implications in different areas, including in complexity theory, additive number theory, information theory, graph theory, and cryptography. Starting from the stronger notion of multicalibration, we obtain stronger and more general versions of a number of applications of the Regularity Lemma, including the Hardcore Lemma, the Dense Model Theorem, and the equivalence of conditional pseudo-min-entropy and unpredictability. For example, we show that every boolean function (regardless of its hardness) has a small collection of disjoint hardcore sets, where the sizes of those hardcore sets are related to how balanced the function is on corresponding pieces of an efficient partition of the domain.

Keywords

Cite

@article{arxiv.2312.17223,
  title  = {Complexity-Theoretic Implications of Multicalibration},
  author = {Sílvia Casacuberta and Cynthia Dwork and Salil Vadhan},
  journal= {arXiv preprint arXiv:2312.17223},
  year   = {2024}
}

Comments

Full version of the paper presented at STOC 2024 and FORC 2024

R2 v1 2026-06-28T14:04:01.176Z