Julien Chevallier
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Julien Chevallier has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Guilherme Ost
Universidade Federal do Rio de Janeiro Rio de Janeiro Brazil
4 shared publications - Anna Melnykova
AGM UMR-CNRS 8088 UMR-CNRS 5224 Université de Cergy-Pontoise Université Grenoble Alpes LJK
3 shared publications - Eva Löcherbach
Statistique, Analyse et Modélisation Multidisciplinaire, Université Paris 1 Panthéon-Sorbonne, EA 4543 et FR FP2M 2036 CNRS
3 shared publications - Irene Tubikanec
Institute of Applied Statistics, Johannes Kepler University Linz
3 shared publications - Maria José Cáceres
Departamento de Matemática Aplicada & IMAG Universidad de Granada Andalusia Spain
2 shared publications Marie Doumic
Inria Paris-Rocquencourt, Domaine de Voluceau, BP 105, 78153 Le Chesnay, France
2 shared publications
Latest Publications
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Modelling large neural networks via Hawkes processes
Modélisation de grands réseaux de neurones par processus de Hawkes: Comment fonctionne le cerveau ? Peut-on créer un cerveau artificiel ? Une étape essentielle en vue d'obtenir une réponse à ces questions est la modélisation mathématique des phénomènes à l'œuvre dans le cerveau. Ce manuscrit se focalise sur l'étude de modèles de réseaux de neurones inspirés de la réalité.Cette thèse se place à la rencontre entre trois grands domaines des mathématiques - l'étude des équations aux dérivées partielles (EDP), les probabilités et la statistique - et s'intéresse à leur application en neurobiologie. Dans un premier temps, nous établissons les liens qui existent entre deux échelles de modélisation neurobiologique. À un niveau microscopique, l'activité électrique de chaque neurone est représentée par un processus ponctuel. À une plus grande échelle, un système d'EDP structuré en âge décrit la dynamique moyenne de ces activités. Il est alors montré que le modèle macroscopique peut se retrouver de deux manières distinctes : en étudiant la dynamique moyenne d'un neurone typique ou bien en étudiant la dynamique d'un réseau de n neurones en champ-moyen quand n tend vers l’infini. Dans le second cas, la convergence vers une dynamique limite est démontrée et les fluctuations de la dynamique microscopique autour de cette limite sont examinées. Dans un second temps, nous construisons une procédure de test d'indépendance entre processus ponctuels, ces derniers étant destinés à modéliser l'activité de certains neurones. Ses performances sont contrôlées théoriquement et vérifiées d'un point de vue pratique par une étude par simulations. Pour finir, notre procédure est appliquée sur de vraies données
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Modelling large neural networks via Hawkes processes
Modélisation de grands réseaux de neurones par processus de Hawkes: Comment fonctionne le cerveau ? Peut-on créer un cerveau artificiel ? Une étape essentielle en vue d'obtenir une réponse à ces questions est la modélisation mathématique des phénomènes à l'œuvre dans le cerveau. Ce manuscrit se focalise sur l'étude de modèles de réseaux de neurones inspirés de la réalité.Cette thèse se place à la rencontre entre trois grands domaines des mathématiques - l'étude des équations aux dérivées partielles (EDP), les probabilités et la statistique - et s'intéresse à leur application en neurobiologie. Dans un premier temps, nous établissons les liens qui existent entre deux échelles de modélisation neurobiologique. À un niveau microscopique, l'activité électrique de chaque neurone est représentée par un processus ponctuel. À une plus grande échelle, un système d'EDP structuré en âge décrit la dynamique moyenne de ces activités. Il est alors montré que le modèle macroscopique peut se retrouver de deux manières distinctes : en étudiant la dynamique moyenne d'un neurone typique ou bien en étudiant la dynamique d'un réseau de n neurones en champ-moyen quand n tend vers l’infini. Dans le second cas, la convergence vers une dynamique limite est démontrée et les fluctuations de la dynamique microscopique autour de cette limite sont examinées. Dans un second temps, nous construisons une procédure de test d'indépendance entre processus ponctuels, ces derniers étant destinés à modéliser l'activité de certains neurones. Ses performances sont contrôlées théoriquement et vérifiées d'un point de vue pratique par une étude par simulations. Pour finir, notre procédure est appliquée sur de vraies données
- 2 April 2026
- Published by Institute of Mathematical Statistics
- Vol. 54 (2) , 861-881
Inferring the dependence graph density of binary graphical models in high dimension
We consider a system of binary interacting chains describing the dynamics of a group of N components that, at each time unit, either send some signal to the others or remain silent otherwise. The interactions among the chains are encoded by a directed Erdös–Rényi random graph with unknown parameter p∈(0,1). Moreover, the system is structured within two populations (excitatory chains versus inhibitory ones), which are coupled via a mean field interaction on the underlying Erdös–Rényi graph. In this paper, we address the question of inferring the connectivity parameter p based only on the observation of the interacting chains over T time units. In our main result, we show that the connectivity parameter p can be estimated with rate N−1/2+N1/2/T+(log(T)/T)1/2 through an easy-to-compute estimator. Our analysis relies on a precise study of the spatiotemporal decay of correlations of the interacting chains. This is done through the study of coalescing random walks defining a backward regeneration representation of the system. Interestingly, we also show that this backward regeneration representation allows us to perfectly sample the system of interacting chains (conditionally on each realization of the underlying Erdös–Rényi graph) from its stationary distribution. These probabilistic results have an interest in its own.
- 2 April 2026
- Published by Institute of Mathematical Statistics
- Vol. 54 (2) , 861-881
Inferring the dependence graph density of binary graphical models in high dimension
We consider a system of binary interacting chains describing the dynamics of a group of N components that, at each time unit, either send some signal to the others or remain silent otherwise. The interactions among the chains are encoded by a directed Erdös–Rényi random graph with unknown parameter p∈(0,1). Moreover, the system is structured within two populations (excitatory chains versus inhibitory ones), which are coupled via a mean field interaction on the underlying Erdös–Rényi graph. In this paper, we address the question of inferring the connectivity parameter p based only on the observation of the interacting chains over T time units. In our main result, we show that the connectivity parameter p can be estimated with rate N−1/2+N1/2/T+(log(T)/T)1/2 through an easy-to-compute estimator. Our analysis relies on a precise study of the spatiotemporal decay of correlations of the interacting chains. This is done through the study of coalescing random walks defining a backward regeneration representation of the system. Interestingly, we also show that this backward regeneration representation allows us to perfectly sample the system of interacting chains (conditionally on each realization of the underlying Erdös–Rényi graph) from its stationary distribution. These probabilistic results have an interest in its own.
- 1 April 2026
- Published by Institute of Mathematical Statistics in The Annals of Statistics
- Vol. 54 (2) , 861-881
Inferring the dependence graph density of binary graphical models in high dimension
We consider a system of binary interacting chains describing the dynamics of a group of N components that, at each time unit, either send some signal to the others or remain silent otherwise. The interactions among the chains are encoded by a directed Erdös–Rényi random graph with unknown parameter p∈(0,1). Moreover, the system is structured within two populations (excitatory chains versus inhibitory ones), which are coupled via a mean field interaction on the underlying Erdös–Rényi graph. In this paper, we address the question of inferring the connectivity parameter p based only on the observation of the interacting chains over T time units. In our main result, we show that the connectivity parameter p can be estimated with rate N−1/2+N1/2/T+(log(T)/T)1/2 through an easy-to-compute estimator. Our analysis relies on a precise study of the spatiotemporal decay of correlations of the interacting chains. This is done through the study of coalescing random walks defining a backward regeneration representation of the system. Interestingly, we also show that this backward regeneration representation allows us to perfectly sample the system of interacting chains (conditionally on each realization of the underlying Erdös–Rényi graph) from its stationary distribution. These probabilistic results have an interest in its own.
- 1 April 2026
- Published by Institute of Mathematical Statistics in The Annals of Statistics
- Vol. 54 (2) , 861-881
Inferring the dependence graph density of binary graphical models in high dimension
We consider a system of binary interacting chains describing the dynamics of a group of N components that, at each time unit, either send some signal to the others or remain silent otherwise. The interactions among the chains are encoded by a directed Erdös–Rényi random graph with unknown parameter p∈(0,1). Moreover, the system is structured within two populations (excitatory chains versus inhibitory ones), which are coupled via a mean field interaction on the underlying Erdös–Rényi graph. In this paper, we address the question of inferring the connectivity parameter p based only on the observation of the interacting chains over T time units. In our main result, we show that the connectivity parameter p can be estimated with rate N−1/2+N1/2/T+(log(T)/T)1/2 through an easy-to-compute estimator. Our analysis relies on a precise study of the spatiotemporal decay of correlations of the interacting chains. This is done through the study of coalescing random walks defining a backward regeneration representation of the system. Interestingly, we also show that this backward regeneration representation allows us to perfectly sample the system of interacting chains (conditionally on each realization of the underlying Erdös–Rényi graph) from its stationary distribution. These probabilistic results have an interest in its own.
- Open Access
Critical point processes obtained from a Gaussian random field with a view toward statistics
This paper establishes the theoretical foundation for statistical applications of an intriguing new type of spatial point processes called critical point processes. These point processes, residing in Euclidean space, consist of the critical points of latent smooth Gaussian random fields or of subsets of critical points like minima, saddle points etc. Despite of the simplicity of their definition, the mathematical analysis of critical point processes is non-trivial involving for example deep results on the geometry of random fields, Sobolev space theory, chaos expansions, and multiple Wiener-Itô integrals. We provide explicit expressions for fundamental moment characteristics used in spatial point process statistics like the intensity parameter, the pair correlation function, and higher order intensity functions. The crucial dependence structure (attraction or repulsiveness) of a critical point process is discussed in depth and is in particular related to the dimension of the points and the type of critical points (extrema, saddle points, or all of the critical points). We propose simulation strategies based on spectral methods or smoothing of grid-based simulations and show that resulting approximate critical point process simulations asymptotically converge to the exact critical point process distribution. Finally, under the increasing domain framework, we obtain asymptotic results for linear and bilinear statistics of a critical point process. In particular, we obtain a multivariate central limit theorem for the intensity parameter estimate and a modified version of Ripley's K-function.
Open AccessCritical point processes obtained from a Gaussian random field with a view toward statistics
This paper establishes the theoretical foundation for statistical applications of an intriguing new type of spatial point processes called critical point processes. These point processes, residing in Euclidean space, consist of the critical points of latent smooth Gaussian random fields or of subsets of critical points like minima, saddle points etc. Despite of the simplicity of their definition, the mathematical analysis of critical point processes is non-trivial involving for example deep results on the geometry of random fields, Sobolev space theory, chaos expansions, and multiple Wiener-Itô integrals. We provide explicit expressions for fundamental moment characteristics used in spatial point process statistics like the intensity parameter, the pair correlation function, and higher order intensity functions. The crucial dependence structure (attraction or repulsiveness) of a critical point process is discussed in depth and is in particular related to the dimension of the points and the type of critical points (extrema, saddle points, or all of the critical points). We propose simulation strategies based on spectral methods or smoothing of grid-based simulations and show that resulting approximate critical point process simulations asymptotically converge to the exact critical point process distribution. Finally, under the increasing domain framework, we obtain asymptotic results for linear and bilinear statistics of a critical point process. In particular, we obtain a multivariate central limit theorem for the intensity parameter estimate and a modified version of Ripley's K-function.
- Open Access
Uniform in time modulus of continuity of Brownian motion
Let $B=(B_t)_{t\geq 0}$ be a standard Brownian motion. The main objective is to find a uniform (in time) control of the modulus of continuity of $B$ in the spirit of what appears in (Kurtz, 1978). More precisely, it involves the control of the exponential moments of the random variable $\sup_{0\leq s\leq t} |B_t-B_s|/w(t,|t-s|)$ for a suitable function $w$. A stability inequality for diffusion processes is then derived and applied to two simple frameworks.Open AccessUniform in time modulus of continuity of Brownian motion
Let $B=(B_t)_{t\geq 0}$ be a standard Brownian motion. The main objective is to find a uniform (in time) control of the modulus of continuity of $B$ in the spirit of what appears in (Kurtz, 1978). More precisely, it involves the control of the exponential moments of the random variable $\sup_{0\leq s\leq t} |B_t-B_s|/w(t,|t-s|)$ for a suitable function $w$. A stability inequality for diffusion processes is then derived and applied to two simple frameworks. - 1 September 2021
- Published by Cambridge University Press (CUP) in Advances in Applied Probability
- Vol. 53 (3) , 716-756
Diffusion approximation of multi-class Hawkes processes: Theoretical and numerical analysis
Oscillatory systems of interacting Hawkes processes with Erlang memory kernels were introduced by Ditlevsen and Löcherbach (Stoch. Process. Appl., 2017). They are piecewise deterministic Markov processes (PDMP) and can be approximated by a stochastic diffusion. In this paper, first, a strong error bound between the PDMP and the diffusion is proved. Second, moment bounds for the resulting diffusion are derived. Third, approximation schemes for the diffusion, based on the numerical splitting approach, are proposed. These schemes are proved to converge with mean-square order 1 and to preserve the properties of the diffusion, in particular the hypoellipticity, the ergodicity, and the moment bounds. Finally, the PDMP and the diffusion are compared through numerical experiments, where the PDMP is simulated with an adapted thinning procedure.
- 1 September 2021
- Published by Cambridge University Press (CUP) in Advances in Applied Probability
- Vol. 53 (3) , 716-756
Diffusion approximation of multi-class Hawkes processes: Theoretical and numerical analysis
Oscillatory systems of interacting Hawkes processes with Erlang memory kernels were introduced by Ditlevsen and Löcherbach (Stoch. Process. Appl., 2017). They are piecewise deterministic Markov processes (PDMP) and can be approximated by a stochastic diffusion. In this paper, first, a strong error bound between the PDMP and the diffusion is proved. Second, moment bounds for the resulting diffusion are derived. Third, approximation schemes for the diffusion, based on the numerical splitting approach, are proposed. These schemes are proved to converge with mean-square order 1 and to preserve the properties of the diffusion, in particular the hypoellipticity, the ergodicity, and the moment bounds. Finally, the PDMP and the diffusion are compared through numerical experiments, where the PDMP is simulated with an adapted thinning procedure.
- Open Access
Natural selection and the advantage of recombination
The ubiquity of sex and recombination in nature is widely viewed as enigmatic, despite an abundance of limited-scope explanations. Natural selection, it seems, should amplify well-matched combinations of genes. Recombination would break up these well-matched combinations and should thus be suppressed. We show, to the contrary, that on average: 1) natural selection amplifies poorly-matched gene combinations and 2) creates time-averaged negative associations in the process. Recombination breaks up these poorly-matched combinations, neutralizes the negative associations, and should thus be passively and universally favored.
Open AccessNatural selection and the advantage of recombination
The ubiquity of sex and recombination in nature is widely viewed as enigmatic, despite an abundance of limited-scope explanations. Natural selection, it seems, should amplify well-matched combinations of genes. Recombination would break up these well-matched combinations and should thus be suppressed. We show, to the contrary, that on average: 1) natural selection amplifies poorly-matched gene combinations and 2) creates time-averaged negative associations in the process. Recombination breaks up these poorly-matched combinations, neutralizes the negative associations, and should thus be passively and universally favored.
- Open Access
Fluctuations for Spatially Extended Hawkes Processes
In a previous paper, it has been shown that the mean-field limit of spatially extended Hawkes processes is characterized as the unique solution $u(t,x)$ of a neural field equation (NFE). The value $u(t,x)$ represents the membrane potential at time $t$ of a typical neuron located in position $x$, embedded in an infinite network of neurons. In the present paper, we complement this result by studying the fluctuations of such a stochastic system around its mean-field limit $u(t,x)$. Our first main result is a central limit theorem stating that the spatial distribution associated with these fluctuations converges to the unique solution of some stochastic differential equation driven by a Gaussian noise. In our second main result, we show that the solutions of this stochastic differential equation can be well approximated by a stochastic version of the neural field equation satisfied by $u(t,x)$. To the best of our knowledge, this result appears to be new in the literature.Open AccessFluctuations for Spatially Extended Hawkes Processes
In a previous paper, it has been shown that the mean-field limit of spatially extended Hawkes processes is characterized as the unique solution $u(t,x)$ of a neural field equation (NFE). The value $u(t,x)$ represents the membrane potential at time $t$ of a typical neuron located in position $x$, embedded in an infinite network of neurons. In the present paper, we complement this result by studying the fluctuations of such a stochastic system around its mean-field limit $u(t,x)$. Our first main result is a central limit theorem stating that the spatial distribution associated with these fluctuations converges to the unique solution of some stochastic differential equation driven by a Gaussian noise. In our second main result, we show that the solutions of this stochastic differential equation can be well approximated by a stochastic version of the neural field equation satisfied by $u(t,x)$. To the best of our knowledge, this result appears to be new in the literature. - Open Access
Theoretical analysis and simulation methods for Hawkes processes and their diffusion approximation
Oscillatory systems of interacting Hawkes processes with Erlang memory kernels were introduced in Ditlevsen (2017). They are piecewise deterministic Markov processes (PDMP) and can be approximated by a stochastic diffusion. First, a strong error bound between the PDMP and the diffusion is proved. Second, moment bounds for the resulting diffusion are derived. Third, approximation schemes for the diffusion, based on the numerical splitting approach, are proposed. These schemes are proved to converge with mean-square order 1 and to preserve the properties of the diffusion, in particular the hypoellipticity, the ergodicity and the moment bounds. Finally, the PDMP and the diffusion are compared through numerical experiments, where the PDMP is simulated with an adapted thinning procedure.
Open AccessTheoretical analysis and simulation methods for Hawkes processes and their diffusion approximation
Oscillatory systems of interacting Hawkes processes with Erlang memory kernels were introduced in Ditlevsen (2017). They are piecewise deterministic Markov processes (PDMP) and can be approximated by a stochastic diffusion. First, a strong error bound between the PDMP and the diffusion is proved. Second, moment bounds for the resulting diffusion are derived. Third, approximation schemes for the diffusion, based on the numerical splitting approach, are proposed. These schemes are proved to converge with mean-square order 1 and to preserve the properties of the diffusion, in particular the hypoellipticity, the ergodicity and the moment bounds. Finally, the PDMP and the diffusion are compared through numerical experiments, where the PDMP is simulated with an adapted thinning procedure.
- Open Access
Theoretical analysis and simulation methods for Hawkes processes and their diffusion approximation
Oscillatory systems of interacting Hawkes processes with Erlang memory kernels were introduced in Ditlevsen and Löcherbach (2017). They are piecewise deterministic Markov processes (PDMP) and can be approximated by a stochastic diffusion. First, a strong error bound between the PDMP and the diffusion is proved. Second, moment bounds for the resulting diffusion are derived. Third, approximation schemes for the diffusion , based on the numerical splitting approach, are proposed. These schemes are proved to converge with mean-square order 1 and to preserve the properties of the diffusion, in particular the hypoellipticity, the ergod-icity and the moment bounds. Finally, the PDMP and the diffusion are compared through numerical experiments, where the PDMP is simulated with an adapted thinning procedure.
Open AccessTheoretical analysis and simulation methods for Hawkes processes and their diffusion approximation
Oscillatory systems of interacting Hawkes processes with Erlang memory kernels were introduced in Ditlevsen and Löcherbach (2017). They are piecewise deterministic Markov processes (PDMP) and can be approximated by a stochastic diffusion. First, a strong error bound between the PDMP and the diffusion is proved. Second, moment bounds for the resulting diffusion are derived. Third, approximation schemes for the diffusion , based on the numerical splitting approach, are proposed. These schemes are proved to converge with mean-square order 1 and to preserve the properties of the diffusion, in particular the hypoellipticity, the ergod-icity and the moment bounds. Finally, the PDMP and the diffusion are compared through numerical experiments, where the PDMP is simulated with an adapted thinning procedure.