OUT-OF-CLASS NOVELTY GENERATION: AN EXPERIMENTAL FOUNDATION

Abstract : Recent advances in machine learning have brought the field closer to computational creativity research. From a creativity research point of view, this offers the potential to study creativity in relationship with knowledge acquisition. From a machine learning perspective, however, several aspects of creativity need to be better defined to allow the machine learning community to develop and test hypotheses in a systematic way. We propose an actionable definition of creativity as the generation of out-of-distribution novelty. We assess several metrics designed for evaluating the quality of generative models on this new task. We also propose a new experimental setup. Inspired by the usual held-out validation, we hold out entire classes for evaluating the generative potential of models. The goal of the novelty generator is then to use training classes to build a model that can generate objects from future (hold-out) classes, unknown at training time-and thus, are novel with respect to the knowledge the model incorporates. Through extensive experiments on various types of generative models, we are able to find architec-tures and hyperparameter combinations which lead to out-of-distribution novelty.
Type de document :
Communication dans un congrès
International Conference on Representation Learning, Apr 2017, Toulon, France
Liste complète des métadonnées

Littérature citée [33 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01773776
Contributeur : Akin Osman Kazakci <>
Soumis le : lundi 23 avril 2018 - 10:35:33
Dernière modification le : lundi 14 mai 2018 - 10:30:38
Document(s) archivé(s) le : mercredi 19 septembre 2018 - 00:02:32

Fichier

out-of-class.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01773776, version 1

Citation

Mehdi Cherti, Balázs Kégl, Akın Kazakçı. OUT-OF-CLASS NOVELTY GENERATION: AN EXPERIMENTAL FOUNDATION. International Conference on Representation Learning, Apr 2017, Toulon, France. 〈hal-01773776〉

Partager

Métriques

Consultations de la notice

80

Téléchargements de fichiers

42