July in Barcelona

Particle and Monte Carlo Methods
Barcelona, July 24-25, 2004

Organizers : Pierre Del Moral, François Le Gland, Éric Moulines, David Nualart and Christian P. Robert

Abstracts, slides, references, etc. are available ! new

The workshop will be held in the Aula Magna room, located in the historical building of Universitat de Barcelona (Gran Via de les Corts Catalanes, 585).

Context

This workshop is a satellite event to the 6th World Congress of the Bernoulli Society and to the 67th Annual Meeting of the Institute of Mathematical Statistics, to be held jointly in Barcelona, July 26-31, 2004.

Support from

CNRS (the French national center for scientific research), through the project (action spécifique du département STIC) AS 67 Méthodes Particulaires,
SMAI (Société de Mathématiques Appliquées & Industrielles),

is gratefully acknowledged.

Objectives

The aim of this two-day workshop is to gather specialists in the domain of particle methods, also known as Monte Carlo methods with interaction, and of other simulation methods, essentially Markov chain Monte-Carlo methods, with applications to

coalescent and genealogical tree models, non linear filtering, statistical inference of hidden Markov models, rare events analysis, mathematical finance, directed polymer simulations, biology, etc.

Scientific background

Intuitively speaking, Monte Carlo methods with interaction are sequential simulation methods, in which particles

explore the state space by mimicking the evolution of an underlying random process,
learn the environment by evaluating a fitness function,
and interact so that only the most successful particles, in view of (the value of) the fitness function, are allowed to survive and to get offsprings at the next generation.

The effect of this mutation / selection mechanism is to automatically concentrate particles (i.e. the available computing power) in regions of interest of the state space. In the special case of particle filtering, which has numerous applications in

positioning, navigation and tracking, visual tracking, mobile robotics, etc.

each particle represents a possible hidden state, and is multiplied or discarded at the next generation on the basis of its consistency with the current observation, as quantified by the likelihood function. In the most general case, particle methods provide approximations of probability distributions associated with a Feynman-Kac flow, by means of the weighted empirical probability distribution associated with an interacting particle system.

Speakers

At the moment, the following speakers have confirmed their participation :

Christophe Andrieu (University of Bristol), Yves Atchadé (Harvard University), Silvia Centanni (Università di Perugia), Rama Cont (CMAP, Palaiseau), Pierre Del Moral (LSP, Toulouse), Persi Diaconis (Stanford University), Susan Holmes (Stanford University), Hans Künsch (ETH, Zürich), François Le Gland (IRISA / INRIA, Rennes), Laurent Miclo (LSP, Toulouse), Éric Moulines (ENST, Paris), Pilar Muñoz (Universitat Politècnica de Catalunya), Anastasia Papavasiliou (Columbia University), Gareth Peters (CUED, Cambridge), Michael Pitt (University of Warwick), Gareth Roberts (Lancaster University), Jeffrey Rosenthal (University of Toronto), Mathias Rousset (LSP, Toulouse), Samy Tindel (IECN, Nancy), Frederi Viens (Purdue University and LAGA, Villetaneuse).

You can find below a list of participants, and a programme.

There are no registration fees : if you are interested in participating to the workshop, please just send us an e-mail.

Particle and Monte Carlo Methods Barcelona, July 24-25, 2004

Particle and Monte Carlo Methods
Barcelona, July 24-25, 2004