Department of Statistics, Oxford

Multivariate point processes are used to model event-type data in a wide range of domains. One interesting application is to model the emission of electric impulses of biological neurons. In this context, the point process model needs to capture the time-dependencies and interactions between neurons, which can be of two kinds: exciting or inhibiting. Estimating these interactions, and in particular the functional connectivity of the neurons are problems that have gained a lot of attention recently. The general nonlinear Hawkes process is a powerful model for events occurring at multiple locations in interaction. Although there is an extensive literature on the analysis of the linear model, the probabilistic and statistical properties of the nonlinear model are still mainly unknown. In this paper, we consider nonlinear Hawkes models and, in a Bayesian nonparametric inference framework, derive concentration rates for the posterior distribution.  We also infer the graph of interactions between the dimensions of the process and prove that the posterior distribution is consistent on the graph adjacency matrix.

We give necessary and sufficient conditions for the variance of the partial sums of stationary processes to be regularly varying in terms of the spectral measure associated with the shift operator. In the case of reversible Markov chains, or with normal transition operator we also give necessary and sufficient conditions in terms of the spectral measure of the transition operator.  

The two spectral measures are then linked through the use of harmonic measure.



This is joint work with S. Utev(University of Leicester, UK) and M. Peligrad (University of Cincinnati, USA).

I will discuss the so-called Lasso method for signal recovery for high-dimensional data and show applications in computational biology, machine learning and image analysis.

Diffusion process models for evolution of neutral genes have a particle dual coalescent process underlying them. Models are reversible with transition functions having a diagonal expansion in orthogonal polynomial eigenfunctions of dimension greater than one, extending classical one-dimensional diffusion models with Beta stationary distribution and Jacobi polynomial expansions to models with Dirichlet or Poisson Dirichlet stationary distributions. Another form of the transition functions is as a mixture depending on the mutant and non-mutant families represented in the leaves of an infinite-leaf coalescent tree.

The one-dimensional Wright-Fisher diffusion process is important in a characterization of a wider class of continuous time reversible Markov processes with Beta stationary distributions originally studied by Bochner (1954) and Gasper (1972). These processes include the subordinated Wright-Fisher diffusion process.

I will take as my starting point a problem which is classical in

population genetics: we wish to understand the distribution of numbers

of individuals in a population who carry different alleles of a

certain gene. We imagine a sample of size n from a population in

which individuals are subject to neutral mutation at a certain

constant rate. Every mutation gives rise to a completely new type.

The genealogy of the sample is modelled by a coalescent process and we

imagine the mutations as a Poisson process of marks along the

coalescent tree. The allelic partition is obtained by tracing back to

the most recent mutation for each individual and grouping together

individuals whose most recent mutations are the same. The number of

blocks of each of the different possible sizes in this partition is

called the allele frequency spectrum. Recently, there has been much

interest in this problem when the underlying coalescent process is a

so-called Lambda-coalescent (even when this is not a biologically

``reasonable'' model) because the allelic partition is a nice example

of an exchangeable random partition. In this talk, I will describe

the asymptotics (as n tends to infinity) of the allele frequency

spectrum when the coalescent process is a particular Lambda-coalescent

which was introduced by Bolthausen and Sznitman. It turns out that

the frequency spectrum scales in a rather unusual way, and that we

need somewhat unusual tools in order to tackle it.

This is joint work with Anne-Laure Basdevant (Toulouse III).

It is well-known that the only space-time scaling limits of Galton-Watson processes are continuous-state branching processes. Their genealogical structure is most explicitly expressed by discrete trees and R-trees, respectively. Weak limit theorems have been recently established for some of these random trees. We study here a Markovian forest growth procedure that allows to construct the genealogical forest of any continuous-state branching process with immigration as an a.s. limit of Galton-Watson forests with edge lengths. Furthermore, we are naturally led to continuous forests with edge lengths. Another strength of our method is that it yields results in the general supercritical case that was excluded in most of the previous literature.

Fragmentation processes model the evolution of a particle that split as time goes on. When small particles split fast enough, the fragmentation is intensive and the initial mass is reduced to dust in finite time. We encode such fragmentation into a continuum random tree (CRT) in the sense of Aldous. When the splitting times are dense near 0, the fragmentation CRT is in turn encoded into a continuous (height) function. Under some mild hypotheses, we calculate the Hausdorff dimension of the CRT, as well as the maximal H