Curvature Beyond Positivity: Greedy Guarantees for Arbitrary Submodular Functions

TL;DR

This paper extends the concept of curvature to all submodular functions, enabling greedy algorithms to provide guarantees even for non-monotone and negative functions, with practical applications demonstrated.

Contribution

It introduces a unified curvature-based analysis for submodular functions that handles negativity and non-monotonicity, providing the first such guarantees beyond classical assumptions.

Findings

Greedy algorithm with pruning achieves curvature-controlled guarantees for all submodular functions.

The new bounds outperform previous guarantees in the non-monotone regime.

Experiments validate the theoretical results on various practical problems.

Abstract

Submodular functions -- functions exhibiting diminishing returns -- are central to machine learning. When the objective is monotone and non-negative, the greedy algorithm achieves a tight $63\%$ approximation. But many practical objectives incorporate costs that make them negative on some inputs, and all existing multiplicative guarantees require non-negativity. Prior work handles negativity through additive bounds for the special class of decomposable functions and non-monotonicity through partial-monotonicity parameters, but these address each difficulty in isolation and neither extends the classical structural theory. We extend \emph{curvature} -- a parameter measuring how far a function deviates from linearity -- to all submodular functions, handling both non-monotonicity and negativity through a single classical concept. A greedy algorithm with pruning achieves a curvature-controlled multiplicative ratio for \emph{any} submodular function, including those taking negative values -- the first such guarantee beyond monotonicity and non-negativity. In the non-monotone regime $1 \le c_g < 2.2$, the bound strictly beats the best known uniform ratio of $0.401$ (for non-negative $f$), and it recovers the classical $(1-e^{-c_g})/c_g$ guarantee for monotone functions. A multilinear-extension variant extends the framework to general combinatorial constraints via multilinear relaxation. Experiments on cost-penalized experimental design, coverage, feature selection, and a curvature sweep on Multi-News passage selection support the theory.