Formal Quality Measures for Predictors in Markov Decision Processes

Christel BaierChristel Baier, Sascha KlüppelholzSascha Klüppelholz, Jakob PiribauerJakob Piribauer, Robin ZiemekRobin Ziemek

Christel Baier, Sascha Klüppelholz, Jakob Piribauer, Robin Ziemek
Formal Quality Measures for Predictors in Markov Decision Processes
Proceedings of the 39th Annual AAAI Conference on Artificial Intelligence, volume 39 of Technical Tracks 25, April 2025. Public Knowledge Project

Details
Bibtex

KurzfassungAbstract
In adaptive systems, predictors are used to anticipate changes in the system’s state or behavior that may require system adaption, e.g., changing its configuration or adjusting resource allocation. Therefore, the quality of predictors is crucial for the overall reliability and performance of the system under control. This paper studies predictors in systems exhibiting probabilistic and non-deterministic behavior modelled as Markov decision processes (MDPs). Main contributions are the introduction of quantitative notions that measure the effectiveness of predictors in terms of their average capability to predict the occurrence of failures or other undesired system behaviors. The average is taken over all memoryless policies. We study two classes of such notions. One class is inspired by concepts that have been introduced in statistical analysis to explain the impact of features on the decisions of binary classifiers (such as precision, recall, f-score). Second, we study a measure that borrows ideas from recent work on probability-raising causality in MDPs and determines the quality of a predictor by the fraction of memoryless policies under which (the set of states in) the predictor is a probability-raising cause for the considered failure scenario.
Weitere Informationen unter:Further Information: Link
Projekt:Project: CPEC
Forschungsgruppe:Research Group: Algebraische und logische Grundlagen der InformatikAlgebraic and Logical Foundations of Computer Science

@article{Baier_Klüppelholz_Piribauer_Ziemek_2025, title={Formal Quality Measures for Predictors in Markov Decision Processes}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/34879}, DOI={10.1609/aaai.v39i25.34879}, number={25}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Baier, Christel and Klüppelholz, Sascha and Piribauer, Jakob and Ziemek, Robin}, year={2025}, month={Apr.}, pages={26760-26768}}