Certified logic-based explainable AI -- the case of monotonic classifiers

DOI10.1007/978-3-031-38828-6_4MaRDI QIDQ6535329zbMATH OpenFDO

Authors Aurélie Hurault, Joao Marques-Silva

Publication date 12 December 2023

zbMATH Keywords

certificationformal explainability

Mathematics Subject Classification ID

Specification and verification (program logics, model checking, etc.) (68Q60) Formalization of mathematics in connection with theorem provers (68V20) General topics in artificial intelligence (68T01)

Recommendations

• Towards formal XAI: formally approximate minimal explanations of neural networks
• A unified logical framework for explanations in classifier systems
• A logic for binary classifiers and their explanation
• Non-monotonic explanation functions
• Feature necessity \& relevancy in ML classifier explanations

Cites work

• A theory of diagnosis from first principles
• Explanation in artificial intelligence: insights from the social sciences
• Formally verifying the solution to the Boolean Pythagorean triples problem
• Handbook of satisfiability. In 2 parts
• Minimal sets on propositional formulae. Problems and reductions

Cited in

(1)

• On the failings of Shapley values for explainability

This page was built for publication: Certified logic-based explainable AI -- the case of monotonic classifiers

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6535329)