← Search

Graziano Mita

2 accepted papers

2021

An Identifiable Double VAE For Disentangled Representations

ICML 2021spotlight

A large part of the literature on learning disentangled representations focuses on variational autoencoders (VAEs). Recent developments demonstrate that disentanglement cannot be obtained in a fully unsupervised setting without inductive biases on models and data. However, Khemakhem et al., AISTATS,…

Cited by 50SourcePDFScholar
2020

LIBRE: Learning Interpretable Boolean Rule Ensembles

AISTATS 2020poster

We present a novel method—LIBRE—learn an interpretable classifier, which materializes as a set of Boolean rules. LIBRE uses an ensemble of bottom-up, weak learners operating on a random subset of features, which allows for the learning of rules that generalize well on unseen data even in imbalanced…

Cited by 25SourcePDFScholar