de Colnet, A., & Marquis, P. (2023). On Translations between ML Models for XAI Purposes. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence (IJCAI-23) (pp. 3158–3166). International Joint Conferences on Artificial Intelligence. https://doi.org/10.24963/ijcai.2023/352
E192-01 - Forschungsbereich Algorithms and Complexity
-
Published in:
Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence (IJCAI-23)
-
ISBN:
978-1-956792-03-4
-
Date (published):
2023
-
Event name:
32nd International Joint Conference on Artificial Intelligence (IJCAI 2023)
en
Event date:
19-Aug-2023 - 25-Aug-2023
-
Event place:
Macao, China
-
Number of Pages:
9
-
Publisher:
International Joint Conferences on Artificial Intelligence
-
Peer reviewed:
Yes
-
Keywords:
Knowledge Representation and Reasoning; Knowledge compilation; Knowledge representation languages
en
Abstract:
In this paper, the succinctness of various ML models is studied. To be more precise, the existence of polynomial-time and polynomial-space translations between representation languages for classifiers is investigated. The languages that are considered include decision trees, random forests, several types of boosted trees, binary neural networks, Boolean multilayer perceptrons, and various logical representations of binary classifiers. We provide a complete map indicating for every pair of languages C, C' whether or not a polynomial-time / polynomial-space translation exists from C to C'. We also explain how to take advantage of the resulting map for XAI purposes.
en
Project title:
Überwindung der Nichthandbarkeit im Knowledge Compilation Map: ESP 235-N (FWF - Österr. Wissenschaftsfonds)
-
Project (external):
French National Research Agency French National Research Agency European Union’s Horizon 2020
