|Location:||Institute of Statistical Mathematics|
|Date:||22nd - 23rd June 2012|
|Title:||Bayesian Inference and Stochastic Computation|
|Aim and Scope:||Recent progress in Bayesian modeling enables us to treat the nonstationary,
inhomogeneous, and non-Gaussian nature of real-world data. Markov chain
Monte Carlo (MCMC), sequential Monte Carlo (SMC), and other stochastic
computational methods play essential roles in this innovative data analysis.
The main aim of this conference is to discuss the recent developments in these fields. However, we want to go beyond this paradigm; our central query is to extend the concept of "stochastic computation based on stochastic modeling."
MCMC and SMC are generic tools for generating high-dimensional random variates and estimating their probabilities; hence, we can apply these methods to other problems of sampling rare events, counting discrete structures, testing, and so on. Thus, it is natural to explore novel applications of MCMC and SMC.
Another area of interest is to unify Bayesian computation to large-scale, multilevel simulation on massively parallel hardware. This approach is sometimes referred to as "data assimilation by stochastic approach."
In this conference, we will collectively explore the key areas of the future of stochastic computation; we hope that it reactivates the Bayesian belief that everything is probabilistic.