Massively Parallel Algorithms for Bayesian Inference and Decision Making
Funding: 2013: $249,234
Project Member(s): Imai, S.
Funding or Partner Organisation: University of New South Wales (The University of New South Wales)
University of Montreal
University of Colorado
Australian Research Council (ARC Discovery Projects)
Start year: 2013
Summary: The primary aims of the project are to develop generic algorithms for econometric and economic analysis suited to graphical processing units, to use them in new and more realistic models of economic behaviour, and to incorporate them in readily accessible software. In accomplishing these aims the project will make fundamental advances in the quantitative and computational infrastructure of economic science and will improve public and private sector decision-making. The principal outcomes of the project will be a reorientation of quantitative economic modeling and statistical inference and higher standards for the realism and reliability of policy recommendations made by economists.
Keywords: Bayesian inference,optimization algorithms
FOR Codes: Mathematical Economics, Economic Framework not elsewhere classified, Econometric and Statistical Methods, Management, Economic Models and Forecasting, Public Sector Productivity