# Research Scientist – Romain Thibaux

Nonparametric Bayesian methods are a generalization of probabilistic models where the prior distribution can be a stochastic process, a distribution over a possibly uncountably infinite number of random variables. The great flexibility of these methods has led to applications in natural language processing, machine vision, computational biology and other fields.

I will introduce Levy processes, which generalize several of these methods. Levy processes are random measures that give independent mass to independent increments. As an example I will show how they can be used to model various types of data such as binary vectors or vectors of counts, with applications to text and images. These techniques are related to the better known Dirichlet process.

**Speaker: Research Scientist - ROMAIN THIBAUX**

*Google Tech Talks*

June, 11 2008

Title: "Levy Processes and Applications to Machine Learning"