Competitive Online Algorithms for Resource Allocation over the Positive Semidefinite Cone

TL;DR

This paper introduces a new online resource allocation framework over the positive semidefinite cone, proposing primal-dual algorithms with optimized competitive ratios for experiment design problems.

Contribution

It develops a convex optimization approach to design smoothing functions satisfying a new PSD diminishing returns property, improving online algorithm performance.

Findings

Derived competitive ratio bounds for primal-dual algorithms.

Optimized smoothing functions for D- and A-optimal experiment design.

Demonstrated improved performance through numerical examples.

Abstract

We consider a new and general online resource allocation problem, where the goal is to maximize a function of a positive semidefinite (PSD) matrix with a scalar budget constraint. The problem data arrives online, and the algorithm needs to make an irrevocable decision at each step. Of particular interest are classic experiment design problems in the online setting, with the algorithm deciding whether to allocate budget to each experiment as new experiments become available sequentially. We analyze two greedy primal-dual algorithms and provide bounds on their competitive ratios. Our analysis relies on a smooth surrogate of the objective function that needs to satisfy a new diminishing returns (PSD-DR) property (that its gradient is order-reversing with respect to the PSD cone). Using the representation for monotone maps on the PSD cone given by L\"owner's theorem, we obtain a convex parametrization of the family of functions satisfying PSD-DR. We then formulate a convex optimization problem to directly optimize our competitive ratio bound over this set. This design problem can be solved offline before the data start arriving. The online algorithm that uses the designed smoothing is tailored to the given cost function, and enjoys a competitive ratio at least as good as our optimized bound. We provide examples of computing the smooth surrogate for D-optimal and A-optimal experiment design, and demonstrate the performance of the custom-designed algorithm.