Online Combinatorial Assignment in Independence Systems

TL;DR

This paper advances online combinatorial assignment algorithms by providing improved competitiveness bounds and polynomial-time solutions for various independence systems, including hypergraph matchings and matroids, in multiple stochastic models.

Contribution

It introduces exponential improvements in competitiveness bounds for online algorithms in independence systems and develops new polynomial-time algorithms using linear programming techniques.

Findings

Upper bounds of O(log(k)/k) and O(log(n)/√n) on competitiveness.

New algorithms with competitive ratios like (1-e^{-k})/k and 1/(k+1).

Polynomial-time algorithms for matroids, matroid intersections, and matchoids.

Abstract

We consider an online multi-weighted generalization of several classic online optimization problems, called the online combinatorial assignment problem. We are given an independence system over a ground set of elements and agents that arrive online one by one. Upon arrival, each agent reveals a weight function over the elements of the ground set. If the independence system is given by the matchings of a hypergraph we recover the combinatorial auction problem, where every node represents an item to be sold, and every edge represents a bundle of items. For combinatorial auctions, Kesselheim et al. showed upper bounds of O(loglog(k)/log(k)) and $O(\log \log(n)/\log(n))$ on the competitiveness of any online algorithm, even in the random order model, where $k$ is the maximum bundle size and $n$ is the number of items. We provide an exponential improvement on these upper bounds to show that the competitiveness of any online algorithm in the prophet IID setting is upper bounded by $O(\log(k)/k)$, and $O(\log(n)/\sqrt{n})$. Furthermore, using linear programming, we provide new and improved guarantees for the $k$-bounded online combinatorial auction problem (i.e., bundles of size at most $k$). We show a $(1-e^{-k})/k$-competitive algorithm in the prophet IID model, a $1/(k+1)$-competitive algorithm in the prophet-secretary model using a single sample per agent, and a $k^{-k/(k-1)}$-competitive algorithm in the secretary model. Our algorithms run in polynomial time and work in more general independence systems where the offline combinatorial assignment problem admits the existence of a polynomial-time randomized algorithm that we call certificate sampler. We show that certificate samplers have a nice interplay with random order models, and we also provide new polynomial-time competitive algorithms for some classes of matroids, matroid intersections, and matchoids.