Learning GAI-decomposable Utility Models for Multiattribute Decision Making - Biologie Computationnelle et Quantitative Access content directly
Conference Papers Year : 2024

Learning GAI-decomposable Utility Models for Multiattribute Decision Making

Margot Herin
  • Function : Author
  • PersonId : 1165630
Patrice Perny


We propose an approach to learn a multiattribute utility function to model, explain or predict the value system of a Decision Maker. The main challenge of the modeling task is to describe human values and preferences in the presence of interacting attributes while keeping the utility function as simple as possible. We focus on the generalized additive decomposable utility model which allows interactions between attributes while preserving some additive decomposability of the evaluation model. We present a learning approach able to identify the factors of interacting attributes and to learn the utility functions defined on these factors. This approach relies on the determination of a sparse representation of the ANOVA decomposition of the multiattribute utility function using multiple kernel learning. It applies to both continuous and discrete attributes. Numerical tests are performed to demonstrate the practical efficiency of the learning approach.
Fichier principal
Vignette du fichier
AAAI24_Preprint.pdf (486.52 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04424705 , version 1 (22-02-2024)


  • HAL Id : hal-04424705 , version 1


Margot Herin, Patrice Perny, Nataliya Sokolovska. Learning GAI-decomposable Utility Models for Multiattribute Decision Making. The 38th Annual AAAI Conference on Artificial Intelligence (AAAI 2024), Feb 2024, Vancouver, Canada. ⟨hal-04424705⟩
111 View
31 Download


Gmail Mastodon Facebook X LinkedIn More