Towards a more efficient computation of individual attribute and policy contribution for post-hoc explanation of cooperative multi-agent systems using Myerson values

Giorgio Angelotti^{a b}✉, Natalia Díaz-Rodríguez^{c d}

1. ^aArtificial and Natural Intelligence Toulouse Institute (ANITI), University of Toulouse, France
2. ^bISAE-Supaero, University of Toulouse, France
3. ^cComputer Science and Artificial Intelligence Department. Andalusian Research Institute in Data Science and Computational Intelligence (DaSCI), Spain
4. ^dResearch Center for Information and Communications Technologies of the University of Granada (CITIC-UGR), Spain
5. ✉Corresponding author