Stéphane Chrétien
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Stéphane Chrétien has a total of 158 co-authors. Here are the top co-authors by number of shared publications:
- Christophe Guyeux
University of Franche-Comte
14 shared publications - Sébastien Darses
IRL CRM–CNRS, Université de Montréal, Centre National de la Recherche Scientifique – Aix-Marseille Université, CNRS, I2M, Marseille, France
6 shared publications - Tianwen Wei
Université de Franche-Comté
5 shared publications - Wenjuan Sun
National Physical Laboratory, Hampton Road, Teddington TW11 0LW, UK
4 shared publications - Astrid Thébault Guiochon
Université Lyon 2
3 shared publications Christophe Guyeux
Université de Franche Comté, IUT Belfort
3 shared publications
Latest Publications
- 17 October 2026
- Published by Springer Nature
- p. 565-582
MultiSigBERT: Beyond Survival Analysis Through Multimodal and Sequential Modeling in Oncology
- 17 October 2026
- Published by Springer Nature
- p. 565-582
MultiSigBERT: Beyond Survival Analysis Through Multimodal and Sequential Modeling in Oncology
- 20 October 2025
- Published by Springer Nature
- p. 83-92
Using Signatures and Koopman Operator to Learn Non-linear Dynamics
- 20 October 2025
- Published by Springer Nature
- p. 83-92
Using Signatures and Koopman Operator to Learn Non-linear Dynamics
- 29 September 2025
- Published by Springer Nature
- p. 391-407
SigBERT: Combining Narrative Medical Reports and Rough Path Signature Theory for Survival Prediction in Oncology
- 29 September 2025
- Published by Springer Nature
- p. 391-407
SigBERT: Combining Narrative Medical Reports and Rough Path Signature Theory for Survival Prediction in Oncology
- Open Access
Conformal Online Learning of Deep Koopman Linear Embeddings
We introduce Conformal Online Learning of Koopman embeddings (COLoKe), a novel framework for adaptively updating Koopman-invariant representations of nonlinear dynamical systems from streaming data. Our modeling approach combines deep feature learning with multi-step prediction consistency in the lifted space, where the dynamics evolve linearly. To prevent overfitting, COLoKe employs a conformal-style mechanism that shifts the focus from evaluating the conformity of new states to assessing the consistency of the current Koopman model. Updates are triggered only when the current model's prediction error exceeds a calibrated threshold, allowing selective refinement of the Koopman operator and embedding. Empirical results on benchmark dynamical systems demonstrate the effectiveness of COLoKe in maintaining long-term predictive accuracy while significantly reducing unnecessary updates. Preprint. Under review.
Open AccessConformal Online Learning of Deep Koopman Linear Embeddings
We introduce Conformal Online Learning of Koopman embeddings (COLoKe), a novel framework for adaptively updating Koopman-invariant representations of nonlinear dynamical systems from streaming data. Our modeling approach combines deep feature learning with multi-step prediction consistency in the lifted space, where the dynamics evolve linearly. To prevent overfitting, COLoKe employs a conformal-style mechanism that shifts the focus from evaluating the conformity of new states to assessing the consistency of the current Koopman model. Updates are triggered only when the current model's prediction error exceeds a calibrated threshold, allowing selective refinement of the Koopman operator and embedding. Empirical results on benchmark dynamical systems demonstrate the effectiveness of COLoKe in maintaining long-term predictive accuracy while significantly reducing unnecessary updates. Preprint. Under review.
- 9 April 2025
- Published by Springer Nature in SN Computer Science
- Vol. 6 (4) , 1-18
Biological Sequence Clustering: Novel Approaches and a Comparative Study
- 9 April 2025
- Published by Springer Nature in SN Computer Science
- Vol. 6 (4) , 1-18
Biological Sequence Clustering: Novel Approaches and a Comparative Study
- Open Access
A Penalized Spline Estimator for Functional Linear Regression with Functional Response
Many scientific studies in recent years have been collecting data at a high frequency, which can be considered as functional data. When both the response variable to be modelled and the covariates are functions, we provide a novel and easy-to-implement method addressing function-on-function linear modelling and obtain interpretable parameters. Two main types of models are considered: the concurrent model which explains the response curve $Y_i(t)$ at time $t$ from the values at same time $t$ of the covariates $X_i^l(t)$; the (feed-forward) integral model which explains $Y_i(t)$ based on the values of covariate curves $X_i^l(s)$ observed at any times $s\leq t$. A regularized inference approach is proposed, which accurately selects an appropriate set of basis functions that can be used for functional data reconstruction and at the same time provides smooth and interpretable functional parameters. A functional confident interval procedure is also proposed which uses the conformalization framework. Numerical studies on simulated data with different scenarios illustrate the good performance our the method to capture the relationship between covariates and response.The method is finally applied to the well-known data in order to compare it to some existed competitors. On Canadian weather data with the problem of predicting precipitations from temperature measurements and on Hawaii ocean data for predicting ocean salinity from temperature, oxygen, chloropigments and density measurements, our method made significant improvements on prediction error.Open AccessA Penalized Spline Estimator for Functional Linear Regression with Functional Response
Many scientific studies in recent years have been collecting data at a high frequency, which can be considered as functional data. When both the response variable to be modelled and the covariates are functions, we provide a novel and easy-to-implement method addressing function-on-function linear modelling and obtain interpretable parameters. Two main types of models are considered: the concurrent model which explains the response curve $Y_i(t)$ at time $t$ from the values at same time $t$ of the covariates $X_i^l(t)$; the (feed-forward) integral model which explains $Y_i(t)$ based on the values of covariate curves $X_i^l(s)$ observed at any times $s\leq t$. A regularized inference approach is proposed, which accurately selects an appropriate set of basis functions that can be used for functional data reconstruction and at the same time provides smooth and interpretable functional parameters. A functional confident interval procedure is also proposed which uses the conformalization framework. Numerical studies on simulated data with different scenarios illustrate the good performance our the method to capture the relationship between covariates and response.The method is finally applied to the well-known data in order to compare it to some existed competitors. On Canadian weather data with the problem of predicting precipitations from temperature measurements and on Hawaii ocean data for predicting ocean salinity from temperature, oxygen, chloropigments and density measurements, our method made significant improvements on prediction error. - 11 July 2024
- Published by Springer Nature in Statistics and Computing
- Vol. 34 (5) , 1-23
A mixture of experts regression model for functional response with functional covariates
- 11 July 2024
- Published by Springer Nature in Statistics and Computing
- Vol. 34 (5) , 1-23
A mixture of experts regression model for functional response with functional covariates
- Open Access
- 18 June 2024
- Published by Springer Nature
- p. C1
Correction to: Leveraging the Power of Signatures for the Construction of Topological Complexes for the Analysis of Multivariate Complex Dynamics
- 18 June 2024
- Published by Springer Nature
- p. C1
Open AccessCorrection to: Leveraging the Power of Signatures for the Construction of Topological Complexes for the Analysis of Multivariate Complex Dynamics
- 14 June 2024
- Published by Springer Nature
- p. 576-587
Computing the Shortest Vector in a Lattice Using the Alternating Direction Method of Multipliers
- 14 June 2024
- Published by Springer Nature
- p. 576-587
Computing the Shortest Vector in a Lattice Using the Alternating Direction Method of Multipliers
- Open Access
Time topological analysis of EEG using signature theory
Anomaly detection in multivariate signals is a task of paramount importance in many disciplines (epidemiology, finance, cognitive sciences and neurosciences, oncology, etc.). In this perspective, Topological Data Analysis (TDA) offers a battery of "shape" invariants that can be exploited for the implementation of an effective detection scheme. Our contribution consists of extending the constructions presented in \cite{chretienleveraging} on the construction of simplicial complexes from the Signatures of signals and their predictive capacities, rather than the use of a generic distance as in \cite{petri2014homological}. Signature theory is a new theme in Machine Learning arXiv:1603.03788 stemming from recent work on the notions of Rough Paths developed by Terry Lyons and his team \cite{lyons2002system} based on the formalism introduced by Chen \cite{chen1957integration}. We explore in particular the detection of changes in topology, based on tracking the evolution of homological persistence and the Betti numbers associated with the complex introduced in \cite{chretienleveraging}. We apply our tools for the analysis of brain signals such as EEG to detect precursor phenomena to epileptic seizures.
Open AccessTime topological analysis of EEG using signature theory
Anomaly detection in multivariate signals is a task of paramount importance in many disciplines (epidemiology, finance, cognitive sciences and neurosciences, oncology, etc.). In this perspective, Topological Data Analysis (TDA) offers a battery of "shape" invariants that can be exploited for the implementation of an effective detection scheme. Our contribution consists of extending the constructions presented in \cite{chretienleveraging} on the construction of simplicial complexes from the Signatures of signals and their predictive capacities, rather than the use of a generic distance as in \cite{petri2014homological}. Signature theory is a new theme in Machine Learning arXiv:1603.03788 stemming from recent work on the notions of Rough Paths developed by Terry Lyons and his team \cite{lyons2002system} based on the formalism introduced by Chen \cite{chen1957integration}. We explore in particular the detection of changes in topology, based on tracking the evolution of homological persistence and the Betti numbers associated with the complex introduced in \cite{chretienleveraging}. We apply our tools for the analysis of brain signals such as EEG to detect precursor phenomena to epileptic seizures.