This paper studies the rate of convergence of an appropriate discretization scheme of the solution of the McKean--Vlasov equation introduced by Bossy and Talay. More specifically, we consider approximations of the distribution and of the density of the solution of the stochastic differential equation associated to the McKean--Vlasov equation. The scheme adopted here is a mixed one: Euler--weakly interacting particle system. If $n$ is the number of weakly interacting particles and $h$ is the uniform step in the time discretization, we prove that the rate of convergence of the distribution functions of the approximating sequence in the $L^1(\Omega\times \Bbb{R})$ norm is of the order of $\frac{1}{\sqrt n} + h $, while for the densities is of the order $ h +\frac{1}{\sqrt {n}h^{1/4}}$. The rates of convergence with respect to the supremum norm are also calculated. This result is obtained by carefully employing techniques of Malliavin calculus.

We discuss a class of anisotropic random walks in a random media on$\mathbb{Z}^{d}$, $d\geq1$, which have reversible transition kernels when the environment is fixed. The aim is to derive a strong law of large numbers and a functional central limit theorem for this class of models. The technique of the environment viewed from the particle does not seem to apply well in this setting. Our approach is based on the technique of introducing certain times similar to the regeneration times in the work concerning random walks in i.i.d. random environment by Sznitman and Zerner. With the help of these times we are able to construct anergodic Markov structure.

We consider the growth rate of a collection of passive tracers moving in the plane under the influence of a random, fluctuating, velocity field. The velocity field we consider is a finite mode approximation to Kolmogorov velocity fields, which are commonly used as models for turbulent diffusion. We show that the diameter of the body of passive tracers grows linearly in time under the influence of these velocity fields.

We prove a shape theorem for a growing set of simple random walks on $\mathbb{Z}^d$, known as the frog model. The dynamics of this process is described as follows: There are active particles, which perform independent discrete time SRWs, and sleeping particles, which do not move. When a sleeping particle is hit by an active particle, it becomes active too. At time $0$ all particles are sleeping, except for that placed at the origin. We prove that the set of the original positions of all active particles, rescaled by the elapsed time, converges to some compact convex set.

The cornerstone of this work, which is partly motivated by the characterization of the so-called eternal additive coalescents by Aldous and Pitman, is an explicit expression for the general eternal solution to Smoluchowski's coagulation equation with additive kernel. This expression points at certain Lévy processes with no negative jumps and more precisely at a stochastic model for aggregation based on such processes, which has been recently considered by Bertoin and Miermont and is known to bear close relations with the additive coalescence. As an application, we show that the eternal solutions can be obtained from some hydrodynamic limit of the stochastic model. We also present a simple condition that ensures the existence of a smooth density for an eternal solution.

We analyze clustering and (local) recurrence of a standard Markov process model of spatial domain coarsening. The continuous time process, whose state space consists of assignments of $+1$ or $-1$ to each site in $\mathbf{Z}^2$, is the zero-temperature limit of the stochastic homogeneous Ising ferromagnet (with Glauber dynamics): the initial state is chosen uniformly at random and then each site, at rate 1, polls its four neighbors and makes sure it agrees with the majority, or tosses a fair coin in case of a tie. Among the main results (almost sure, with respect to both the process and initial state) are: clusters (maximal domains of constant sign) are finite for times $t< \infty$, but the cluster of a fixed site diverges (in diameter) as $t \to \infty$; each of the two constant states is (positive) recurrent. We also present other results and conjectures concerning positive and null recurrence and the role of absorbing states.

In this paper we develop tools for analyzing the rate at which a reversible Markov chain converges to stationarity. Our techniques are useful when the Markov chain can be decomposed into pieces which are themselves easier to analyze. The main theorems relate the spectral gap of the original Markov chains to the spectral gaps of the pieces. In the first case the pieces are restrictions of the Markov chain to subsets of the state space; the second case treats a Metropolis--Hastings chain whose equilibrium distribution is a weighted average of equilibrium distributions of other Metropolis--Hastings chains on the same state space.

In the random allocation model, balls are sequentially inserted at random into $n$ exchangeable bins. The occupancy score of a bin denotes the number of balls inserted in this bin. The (random) distribution of occupancy scores defines the object of this paper: the empirical occupancy measure which is a probability measure over the integers. This measure-valued random variable packages many useful statistics. This paper characterizes the large deviations of the flow of empirical occupancy measures when $n$ goes to infinity while the number of inserted balls remains proportional to $n$. The main result is a Sanov-like theorem for the empirical occupancy measure when the set of probability measures over the integers is endowed with metrics that are slightly stronger than the total variation distance. Thanks to a coupling argument, this result applies to the degree distribution of sparse random graphs.

We derive a family of approximate sampling distributions for the symmetric overdominance model of population genetics. The distributions are selective versions of the Ewens Sampling Formula, which gives sample likelihoods under a model of neutral evolution. We draw on basic results for the general selection model of Ethier and Kurtz, and use mathematical tools well-suited for calculating expectations of symmetric functions of Poisson--Dirichlet atoms. We conclude by briefly examining a Human Leukocyte Antigen data set, in light of a distribution conditional on the number of sample atoms.

We analyze numerical methods for the pathwise approximation of a system of stochastic differential equations. As a measure of performance we consider the $q$th mean of the maximum distance between the solution and its approximation on the whole unit interval. We introduce an adaptive discretization that takes into account the local smoothness of every trajectory of the solution. The resulting adaptive Euler approximation performs asymptotically optimal in the class of all numerical methods that are based on a finite number of observations of the driving Brownian motion.

We provide the solution to a fusion of two fundamental problems in mathematical finance. The first problem is that of maximizing the expected utility of terminal wealth of an investor who holds a short position in a contingent claim, and the second is that of maximizing terminal wealth where the utility function allows the investor to have negative wealth. Under assumptions of reasonable asymptotic elasticity on the investor's utility function, we present an optimal investment theorem and simultaneously treat the corresponding dual problem.

In this article we discuss Markovian term structure models in discrete time and with continuous state space. More precisely, we are concerned with the structural properties of such models if one has the Markov property for a part of the forward curve. We investigate the two cases where these parts are either a true subset of the forward curve, including the short rate, or the entire forward curve. For the former case we give a sufficient condition for the term structure model to be affine. For the latter case we provide a version of the Heath, Jarrow and Morton drift condition. Under a Gaussian assumption a Heath--Jarrow--Morton--Musiela type equation is derived.

We consider an optimal investment problem proposed by Bielecki and Pliska. The goal of the investment problem is to optimize the long-term growth of expected utility of wealth. We consider HARA utility functions with exponent $-\infty< \gamma< 1$. The problem can be reformulated as an infinite time horizon risk-sensitive control problem. Some useful ideas and results from the theory of risk-sensitive control can be used in the analysis. Especially, we analyze the associated dynamical programming equation. Then an optimal (or approximately optimal) Markovian investment policy can be derived.

Consider a Gaussian random field with a finite Karhunen--Loève expansion of the form $Z(u) = \sum_{i=1}^n u_i z_i$, where $z_i$, $i=1,\ldots,n,$ are independent standard normal variables and $u=(u_1,\ldots,u_n)'$ ranges over an index set $M$, which is a subset of the unit sphere $S^{n-1}$ in $R^n$. Under a very general assumption that $M$ is a manifold with a piecewise smooth boundary, we prove the validity and the equivalence of two currently available methods for obtaining the asymptotic expansion of the tail probability of the maximum of $Z(u)$. One is the tube method, where the volume of the tube around the index set $M$ is evaluated. The other is the Euler characteristic method, where the expectation for the Euler characteristic of the excursion set is evaluated. General discussion on this equivalence was given in a recent paper by R. J. Adler. In order to show the equivalence we prove a version of the Morse theorem for a manifold with a piecewise smooth boundary.