Skip to Main content Skip to Navigation
Conference papers

ICE-BeeM: Identifiable Conditional Energy-Based Deep Models Based on Nonlinear ICA

Abstract : We consider the identifiability theory of probabilistic models and establish sufficient conditions under which the representations learned by a very broad family of conditional energy-based models are unique in function space, up to a simple transformation. In our model family, the energy function is the dot-product between two feature extractors, one for the dependent variable, and one for the conditioning variable. We show that under mild conditions, the features are unique up to scaling and permutation. Our results extend recent developments in nonlinear ICA, and in fact, they lead to an important generalization of ICA models. In particular, we show that our model can be used for the estimation of the components in the framework of Independently Modulated Component Analysis (IMCA), a new generalization of nonlinear ICA that relaxes the independence assumption. A thorough empirical study shows that representations learned by our model from real-world image datasets are identifiable, and improve performance in transfer learning and semi-supervised learning tasks.
Document type :
Conference papers
Complete list of metadata
Contributor : Aapo Hyvärinen Connect in order to contact the contributor
Submitted on : Tuesday, December 1, 2020 - 4:23:58 PM
Last modification on : Monday, December 13, 2021 - 9:16:11 AM

Links full text


  • HAL Id : hal-03034169, version 1
  • ARXIV : 2002.11537


Ilyes Khemakhem, Ricardo Pio Monti, Diederik Kingma, Aapo Hyvärinen. ICE-BeeM: Identifiable Conditional Energy-Based Deep Models Based on Nonlinear ICA. NeurIPS 2020 - 34th Conference on Neural Information Processing Systems, Dec 2020, Vancouver / Virtuel, Canada. ⟨hal-03034169⟩



Les métriques sont temporairement indisponibles