Online Algorithms with Unreliable Guidance
#Online Algorithms #Unreliable Guidance #Machine Learning #Decision Making #DTB Compiler #Consistency-Robustness #arXiv
📌 Key Takeaways
- Researchers introduced a new OAG model for ML-augmented online decision making
- The DTB compiler transforms online algorithms into learning-augmented versions
- The approach provides optimal solutions for caching and metrical task systems
- The algorithm outperforms state-of-the-art for bipartite matching problems
📖 Full Retelling
Computer scientists Julien Dallot, Yuval Emek, Yuval Gil, Maciej Pacut, and Stefan Schmid introduced a new model for machine learning-augmented online decision making called 'online algorithms with unreliable guidance' (OAG) in their paper submitted to arXiv on February 24, 2026, aiming to create a unified framework that separates predictive and algorithmic components in online decision systems. The OAG model, formulated through the lens of request-answer games, provides algorithms with guidance for each incoming request that ideally would be the optimal answer, but with probability β, the guidance is adversarially corrupted. The researchers developed this model to address the challenge of creating algorithms that perform well in both ideal conditions (when β = 0, known as consistency) and when guidance is completely unreliable (when β = 1, known as robustness). To achieve this, they introduced a systematic method called the 'drop or trust blindly compiler' that transforms any online algorithm into a learning-augmented version by making it either follow incoming guidance blindly or ignore it completely based on a coin toss outcome. Despite its simplicity, this approach produces algorithms with attractive consistency-robustness guarantees for three classic online problems: for caching and uniform metrical task systems, their algorithms are optimal, while for bipartite matching with adversarial arrival order, their algorithm outperforms existing state-of-the-art solutions, representing a significant advancement in the field of online algorithms with machine learning augmentation.
🏷️ Themes
Machine Learning, Online Algorithms, Decision Making
📚 Related People & Topics
Decision-making
Process to choose a course of action
In psychology, decision-making (also spelled decision making and decisionmaking) is regarded as the cognitive process resulting in the selection of a belief or a course of action among several possible alternative options. It could be either rational or irrational. The decision-making process is a r...
Machine learning
Study of algorithms that improve automatically through experience
Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of statistical algorithms that can learn from data and generalize to unseen data, and thus perform tasks without explicit instructions. Within a subdiscipline in machine learning, advances i...
Entity Intersection Graph
No entity connections available yet for this article.
Mentioned Entities
Original Source
--> Computer Science > Artificial Intelligence arXiv:2602.20706 [Submitted on 24 Feb 2026] Title: Online Algorithms with Unreliable Guidance Authors: Julien Dallot , Yuval Emek , Yuval Gil , Maciej Pacut , Stefan Schmid View a PDF of the paper titled Online Algorithms with Unreliable Guidance, by Julien Dallot and 3 other authors View PDF HTML Abstract: This paper introduces a new model for ML-augmented online decision making, called online algorithms with unreliable guidance . This model completely separates between the predictive and algorithmic components, thus offering a single well-defined analysis framework that relies solely on the considered problem. Formulated through the lens of request-answer games, an OAG algorithm receives, with each incoming request, a piece of guidance which is taken from the problem's answer space; ideally, this guidance is the optimal answer for the current request, however with probability $\beta$, the guidance is adversarially corrupted. The goal is to develop OAG algorithms that admit good competitiveness when $\beta = 0$ (a.k.a. consistency) as well as when $\beta = 1$ (a.k.a. robustness); the appealing notion of smoothness, that in most prior work required a dedicated loss function, now arises naturally as $\beta$ shifts from $0$ to $1$. We then describe a systematic method, called the drop or trust blindly compiler, which transforms any online algorithm into a learning-augmented online algorithm in the OAG model. Given a prediction-oblivious online algorithm, its learning-augmented counterpart produced by applying the DTB compiler either follows the incoming guidance blindly or ignores it altogether and proceeds as the initial algorithm would have; the choice between these two alternatives is based on the outcome of a coin toss. As our main technical contribution, we prove that although remarkably simple, the class of algorithms produced via the DTB compiler includes algorithms with attractive consistency-robustness guarantees...
Read full article at source