Exact approximate Monte Carlo methods for complex Bayesian inference (research costs) (EAMC)


Main funder


Funds granted by main funder (€)

140 000,00


Funding program

Research costs of Academy Research Fellow, AoF (Academy of Finland)


Project timetable

Project start date: 01/09/2017

Project end date: 31/08/2019


Summary

Bayesian statistics is a flexible and coherent framework to model different types of phenomena involving uncertainty. The applicability of Bayesian methods is often constrained by computational challenges, which can often be expressed in terms of high-dimensional and difficult integrals. Such situations may occur for example in time-series analysis, inverse problems and design of experiments.

The project focuses on an emerging class of computational methods, which are more efficient than earlier methods in complex situations, such as the above mentioned application domains. The research questions span from theoretical and methodological questions to applications. If succesful, the outcome of the project is research findings that lead into new more efficient computational methods, and open doors for new types of more challenging applications for Bayesian statistics.


Principal Investigator


Primary responsible unit


Related publications


Last updated on 2019-08-10 at 11:14