The Primal-Dual method for Learning Augmented Algorithms

TL;DR

This paper extends the primal-dual method to incorporate predictions in online algorithms, resulting in new algorithms for covering problems that perform well with accurate predictions and remain robust when predictions are inaccurate.

Contribution

It introduces a primal-dual framework for learning-augmented online algorithms, improving performance by leveraging predictions for a range of covering problems.

Findings

Algorithms outperform traditional online methods with accurate predictions

Maintain strong guarantees even when predictions are misleading

Applicable to various online covering problems

Abstract

The extension of classical online algorithms when provided with predictions is a new and active research area. In this paper, we extend the primal-dual method for online algorithms in order to incorporate predictions that advise the online algorithm about the next action to take. We use this framework to obtain novel algorithms for a variety of online covering problems. We compare our algorithms to the cost of the true and predicted offline optimal solutions and show that these algorithms outperform any online algorithm when the prediction is accurate while maintaining good guarantees when the prediction is misleading.