no code implementations • 14 Jan 2025 • Marine Hamon, Vincent Lemaire, Nour Eddine Yassine Nair-Benrekia, Samuel Berlemont, Julien Cumin
To detect anomalies with precision and without prior knowledge in time series, is it better to build a detector from the initial temporal representation, or to compute a new (tabular) representation using an existing automatic variable construction library?
no code implementations • 19 Nov 2024 • Vincent Lemaire, Gilles Pagès, Christian Yeo
We introduce a new class of neural networks designed to be convex functions of their inputs, leveraging the principle that any convex function can be represented as the supremum of the affine functions it dominates.
1 code implementation • 21 Oct 2024 • Thomas George, Pierre Nodet, Alexis Bondu, Vincent Lemaire
Mislabeled examples are ubiquitous in real-world machine learning datasets, advocating the development of techniques for automatic detection.
no code implementations • 12 Sep 2024 • Arthur Hoarau, Vincent Lemaire
Recent research in machine learning has given rise to a flourishing literature on the quantification and decomposition of model uncertainty.
2 code implementations • 23 Aug 2024 • Aurélien Renault, Youssef Achenchabe, Édouard Bertrand, Alexis Bondu, Antoine Cornuéjols, Vincent Lemaire, Asma Dachraoui
\texttt{ml\_edm} is a Python 3 library, designed for early decision making of any learning tasks involving temporal/sequential data.
2 code implementations • 26 Jun 2024 • Aurélien Renault, Alexis Bondu, Antoine Cornuéjols, Vincent Lemaire
In many situations, the measurements of a studied phenomenon are provided sequentially, and the prediction of its class needs to be made as early as possible so as not to incur too high a time penalty, but not too early and risk paying the cost of misclassification.
1 code implementation • 11 Mar 2024 • Colin Troisemaine, Vincent Lemaire
This paper proposes a method for the automatic creation of variables (in the case of regression) that complement the information contained in the initial input vector.
no code implementations • 7 Feb 2024 • Stanislas Strasman, Antonio Ocello, Claire Boyer, Sylvain Le Corff, Vincent Lemaire
Score-based generative models (SGMs) aim at estimating a target data distribution by learning score functions using only noise-perturbed samples from the target. Recent literature has focused extensively on assessing the error between the target and estimated distributions, gauging the generative quality through the Kullback-Leibler (KL) divergence and Wasserstein distances.
2 code implementations • 9 Nov 2023 • Colin Troisemaine, Alexandre Reiffers-Masson, Stéphane Gosselin, Vincent Lemaire, Sandrine Vaton
In particular, the number of novel classes is usually assumed to be known in advance, and their labels are sometimes used to tune hyperparameters.
no code implementations • 21 Sep 2023 • Arthur Hoarau, Vincent Lemaire, Arnaud Martin, Jean-Christophe Dubois, Yolande Le Gall
Recent research in active learning, and more precisely in uncertainty sampling, has focused on the decomposition of model uncertainty into reducible and irreducible uncertainties.
1 code implementation • Advanced Analytics and Learning on Temporal Data 2023 • Arik Ermshaus, Patrick Schäfer, Anthony Bagnall, Thomas Guyet, Georgiana Ifrim, Vincent Lemaire, Ulf Leser, Colin Leverger, Simon Malinowski
Despite its importance, existing methods demonstrate limited efficacy on real-world multivariate time series data.
no code implementations • 8 Sep 2023 • Vincent Lemaire, Nathan Le Boudec, Victor Guyomard, Françoise Fessant
There are now many explainable AI methods for understanding the decisions of a machine learning model.
1 code implementation • 29 Aug 2023 • Pierre Nodet, Vincent Lemaire, Alexis Bondu, Antoine Cornuéjols
Training machine learning models from data with weak supervision and dataset shifts is still challenging.
1 code implementation • 18 Aug 2023 • Pierre Nodet, Vincent Lemaire, Alexis Bondu, Antoine Cornuéjols
That is why Biquality Learning has been proposed as a machine learning framework to design algorithms capable of handling multiple weaknesses of supervision and dataset shifts without assumptions on their nature and level by relying on the availability of a small trusted dataset composed of cleanly labeled and representative samples.
no code implementations • 2 Aug 2023 • Aurélien Renault, Alexis Bondu, Vincent Lemaire, Dominique Gay
Time Series Classification (TSC) has received much attention in the past two decades and is still a crucial and challenging problem in data science and knowledge engineering.
no code implementations • 31 Jul 2023 • Vincent Lemaire, Fabrice Clérot, Marc Boullé
In the case of the naive Bayes classifier, and to our knowledge, there is no ``analytical" formulation of Shapley values.
1 code implementation • 22 Jun 2023 • Colin Troisemaine, Joachim Flocon-Cholet, Stéphane Gosselin, Alexandre Reiffers-Masson, Sandrine Vaton, Vincent Lemaire
This task is difficult and can often only be performed by a domain expert.
1 code implementation • 6 Jun 2023 • Vincent Lemaire, Gilles Pagès, Christian Yeo
We propose two parametric approaches to evaluate swing contracts with firm constraints.
no code implementations • 10 Mar 2023 • Daphné Giorgi, Sarah Kaakai, Vincent Lemaire
The R Package IBMPopSim aims to simulate the random evolution of heterogeneous populations using stochastic Individual-Based Models (IBMs).
2 code implementations • 22 Feb 2023 • Colin Troisemaine, Vincent Lemaire, Stéphane Gosselin, Alexandre Reiffers-Masson, Joachim Flocon-Cholet, Sandrine Vaton
We then give an overview of the different families of approaches, organized by the way they transfer knowledge from the labeled set to the unlabeled set.
1 code implementation • 28 Nov 2022 • Colin Troisemaine, Joachim Flocon-Cholet, Stéphane Gosselin, Sandrine Vaton, Alexandre Reiffers-Masson, Vincent Lemaire
In Novel Class Discovery (NCD), the goal is to find new classes in an unlabeled set given a labeled set of known but different classes.
2 code implementations • 2 Sep 2022 • Colin Troisemaine, Joachim Flocon-Cholet, Stéphane Gosselin, Sandrine Vaton, Alexandre Reiffers-Masson, Vincent Lemaire
In this paper, we propose TabularNCD, a new method for discovering novel classes in tabular data.
1 code implementation • 27 Apr 2022 • Alexis Bondu, Youssef Achenchabe, Albert Bifet, Fabrice Clérot, Antoine Cornuéjols, Joao Gama, Georges Hébrail, Vincent Lemaire, Pierre-François Marteau
However, the later a decision is made, the more its accuracy tends to improve, since the description of the problem to hand is enriched over time.
no code implementations • 1 Apr 2022 • Youssef Achenchabe, Alexis Bondu, Antoine Cornuéjols, Vincent Lemaire
In the Early Classification in Open Time Series problem (ECOTS), the task is to predict events, i. e. their class and time interval, at the moment that optimizes the accuracy vs. earliness trade-off.
1 code implementation • 3 Dec 2021 • Colin Troisemaine, Vincent Lemaire
This paper proposes a method for the automatic creation of variables (in the case of regression) that complement the information contained in the initial input vector.
no code implementations • 21 Sep 2021 • Youssef Achenchabe, Alexis Bondu, Antoine Cornuéjols, Vincent Lemaire
Many approaches have been proposed for early classification of time series in light of itssignificance in a wide range of applications including healthcare, transportation and fi-nance.
no code implementations • 20 Aug 2021 • Pierre Nodet, Vincent Lemaire, Alexis Bondu, Antoine Cornuéjols
In presence of noise the experiments show that fine tuning of Contrastive representation allows the six methods to achieve better results than end-to-end learning and represent a new reference compare to the recent state of art.
no code implementations • 27 Apr 2021 • Youssef Achenchabe, Alexis Bondu, Antoine Cornuéjols, Vincent Lemaire
Many approaches have been proposed for early classification of time series in light of its significance in a wide range of applications including healthcare, transportation and finance.
no code implementations • 15 Mar 2021 • Dominique Gay, Alexis Bondu, Vincent Lemaire, Marc Boullé
Supervised learning of time series data has been extensively studied for the case of a categorical target variable.
no code implementations • 16 Dec 2020 • Vincent Lemaire, Oumaima Alaoui Ismaili, Antoine Cornuéjols, Dominique Gay
We present two new algorithms using this technique and show on a variety of data sets that they are competitive for prediction performance with pure supervised classifiers while offering interpretability of the clusters discovered.
no code implementations • 16 Dec 2020 • Louis Desreumaux, Vincent Lemaire
To this end, we present the results of a benchmark performed on 20 datasets that compares a strategy learned using a recent meta-learning algorithm with margin sampling.
no code implementations • 16 Dec 2020 • Pierre Nodet, Vincent Lemaire, Alexis Bondu, Antoine Cornuéjols, Adam Ouorou
The field of Weakly Supervised Learning (WSL) has recently seen a surge of popularity, with numerous papers addressing different types of "supervision deficiencies".
no code implementations • 22 Oct 2020 • Hugo Le Baher, Vincent Lemaire, Romain Trinquart
In some industrial applications such as fraud detection, the performance of common supervision techniques may be affected by the poor quality of the available labels : in actual operational use-cases, these labels may be weak in quantity, quality or trustworthiness.
1 code implementation • 19 Oct 2020 • Pierre Nodet, Vincent Lemaire, Alexis Bondu, Antoine Cornuéjols
The field of Weakly Supervised Learning (WSL) has recently seen a surge of popularity, with numerous papers addressing different types of "supervision deficiencies", namely: poor quality, non adaptability, and insufficient quantity of labels.
no code implementations • 13 Nov 2019 • Jean-Michel Fayolle, Vincent Lemaire, Thibaut Montes, Gilles Pagès
This paper proposes two numerical solution based on Product Optimal Quantization for the pricing of Foreign Echange (FX) linked long term Bermudan options e. g. Bermudan Power Reverse Dual Currency options, where we take into account stochastic domestic and foreign interest rates on top of stochastic FX rate, hence we consider a 3-factor model.
no code implementations • WS 2019 • Jean-Leon Bouraoui, Sonia Le Meitour, Romain Carbou, Lina M. Rojas Barahona, Vincent Lemaire
We present Graph2Bots, a tool for assisting conversational agent designers.
no code implementations • 8 Mar 2019 • Dominique Gay, Vincent Lemaire
Since the introduction and the public availability of the \textsc{ucr} time series benchmark data sets, numerous Time Series Classification (TSC) methods has been designed, evaluated and compared to each others.
no code implementations • 6 Nov 2018 • Colin Leverger, Vincent Lemaire, Simon Malinowski, Thomas Guyet, Laurence Rozé
In the context of capacity planning, forecasting the evolution of informatics servers usage enables companies to better manage their computational resources.