Search Results for author: Pierre-Alexandre Mattei

Found 21 papers, 5 papers with code

Are Ensembles Getting Better all the Time?

1 code implementation29 Nov 2023 Pierre-Alexandre Mattei, Damien Garreau

More precisely, in that case, the average loss of the ensemble is a decreasing function of the number of models.

Generalised Mutual Information: a Framework for Discriminative Clustering

no code implementations6 Sep 2023 Louis Ohl, Pierre-Alexandre Mattei, Charles Bouveyron, Warith Harchaoui, Mickaël Leclercq, Arnaud Droit, Frédéric Precioso

In the last decade, recent successes in deep clustering majorly involved the Mutual Information (MI) as an unsupervised objective for training neural networks with increasing regularisations.

Clustering Deep Clustering

Fed-MIWAE: Federated Imputation of Incomplete Data via Deep Generative Models

no code implementations17 Apr 2023 Irene Balelli, Aude Sportisse, Francesco Cremonesi, Pierre-Alexandre Mattei, Marco Lorenzi

In addition, thanks to the variational nature of Fed-MIWAE, our method is designed to perform multiple imputation, allowing for the quantification of the imputation uncertainty in the federated scenario.

Federated Learning Imputation

Are labels informative in semi-supervised learning? -- Estimating and leveraging the missing-data mechanism

no code implementations15 Feb 2023 Aude Sportisse, Hugo Schmutz, Olivier Humbert, Charles Bouveyron, Pierre-Alexandre Mattei

Semi-supervised learning is a powerful technique for leveraging unlabeled data to improve machine learning models, but it can be affected by the presence of ``informative'' labels, which occur when some classes are more likely to be labeled than others.

Data Augmentation

Sparse GEMINI for Joint Discriminative Clustering and Feature Selection

no code implementations7 Feb 2023 Louis Ohl, Pierre-Alexandre Mattei, Charles Bouveyron, Mickaël Leclercq, Arnaud Droit, Frédéric Precioso

Feature selection in clustering is a hard task which involves simultaneously the discovery of relevant clusters as well as relevant variables with respect to these clusters.

Clustering feature selection +1

Explainability as statistical inference

no code implementations6 Dec 2022 Hugo Henri Joseph Senetaire, Damien Garreau, Jes Frellsen, Pierre-Alexandre Mattei

The model parameters can be learned via maximum likelihood, and the method can be adapted to any predictor network architecture and any type of prediction problem.

Imputation

Generalised Mutual Information for Discriminative Clustering

1 code implementation12 Oct 2022 Louis Ohl, Pierre-Alexandre Mattei, Charles Bouveyron, Warith Harchaoui, Mickaël Leclercq, Arnaud Droit, Frederic Precioso

In the last decade, recent successes in deep clustering majorly involved the mutual information (MI) as an unsupervised objective for training neural networks with increasing regularisations.

Clustering Deep Clustering

A Multi-stage deep architecture for summary generation of soccer videos

no code implementations2 May 2022 Melissa Sanabria, Frédéric Precioso, Pierre-Alexandre Mattei, Thomas Menguy

The results show that our method can detect the actions of the match, identify which of these actions should belong to the summary and then propose multiple candidate summaries which are similar enough but with relevant variability to provide different options to the final editor.

Sports Analytics Video Summarization

Model-agnostic out-of-distribution detection using combined statistical tests

no code implementations2 Mar 2022 Federico Bergamin, Pierre-Alexandre Mattei, Jakob D. Havtorn, Hugo Senetaire, Hugo Schmutz, Lars Maaløe, Søren Hauberg, Jes Frellsen

These techniques, based on classical statistical tests, are model-agnostic in the sense that they can be applied to any differentiable generative model.

Out-of-Distribution Detection

Uphill Roads to Variational Tightness: Monotonicity and Monte Carlo Objectives

no code implementations26 Jan 2022 Pierre-Alexandre Mattei, Jes Frellsen

Inspired by this simple monotonicity theorem, we present a series of nonasymptotic results that link properties of Monte Carlo estimates to tightness of MCOs.

Variational Inference

How to deal with missing data in supervised deep learning?

no code implementations ICLR 2022 Niels Bruun Ipsen, Pierre-Alexandre Mattei, Jes Frellsen

To address supervised deep learning with missing values, we propose to marginalize over missing values in a joint model of covariates and outcomes.

Inductive Bias Variational Inference

Tensor decomposition for learning Gaussian mixtures from moments

no code implementations1 Jun 2021 Rima Khouja, Pierre-Alexandre Mattei, Bernard Mourrain

In data processing and machine learning, an important challenge is to recover and exploit models that can represent accurately the data.

Tensor Decomposition

Unobserved classes and extra variables in high-dimensional discriminant analysis

no code implementations3 Feb 2021 Michael Fop, Pierre-Alexandre Mattei, Charles Bouveyron, Thomas Brendan Murphy

In supervised classification problems, the test set may contain data points belonging to classes not observed in the learning phase.

Classification General Classification +2

not-MIWAE: Deep Generative Modelling with Missing not at Random Data

1 code implementation ICLR 2021 Niels Bruun Ipsen, Pierre-Alexandre Mattei, Jes Frellsen

When a missing process depends on the missing values themselves, it needs to be explicitly modelled and taken into account while doing likelihood-based inference.

Variational Inference

Partially Exchangeable Networks and Architectures for Learning Summary Statistics in Approximate Bayesian Computation

1 code implementation29 Jan 2019 Samuel Wiqvist, Pierre-Alexandre Mattei, Umberto Picchini, Jes Frellsen

We present a novel family of deep neural architectures, named partially exchangeable networks (PENs) that leverage probabilistic symmetries.

Time Series Time Series Analysis

MIWAE: Deep Generative Modelling and Imputation of Incomplete Data

no code implementations6 Dec 2018 Pierre-Alexandre Mattei, Jes Frellsen

Our approach, called MIWAE, is based on the importance-weighted autoencoder (IWAE), and maximises a potentially tight lower bound of the log-likelihood of the observed data.

Imputation

Leveraging the Exact Likelihood of Deep Latent Variable Models

no code implementations NeurIPS 2018 Pierre-Alexandre Mattei, Jes Frellsen

Finally, we describe an algorithm for missing data imputation using the exact conditional likelihood of a deep latent variable model.

Imputation

Exact Dimensionality Selection for Bayesian PCA

no code implementations8 Mar 2017 Charles Bouveyron, Pierre Latouche, Pierre-Alexandre Mattei

We present a Bayesian model selection approach to estimate the intrinsic dimensionality of a high-dimensional dataset.

Model Selection

Bayesian Variable Selection for Globally Sparse Probabilistic PCA

no code implementations19 May 2016 Charles Bouveyron, Pierre Latouche, Pierre-Alexandre Mattei

To this end, using Roweis' probabilistic interpretation of PCA and a Gaussian prior on the loading matrix, we provide the first exact computation of the marginal likelihood of a Bayesian PCA model.

Model Selection Variable Selection

Cannot find the paper you are looking for? You can Submit a new open access paper.