Discovery Early Career Researcher Award - Grant ID: DE160100999
Funder
Australian Research Council
Funding Amount
$295,020.00
Summary
Applying forward-backward stochastic differential equations to optimisation. This project intends to develop new ways to solve optimisation problems that are currently difficult to solve because of their complexity and size. In particular, forward–backward stochastic differential equations (FBSDEs) are a new technique that is showing ways to solve problems for which there is yet to be a solution. This project's focus will be on problems that cannot use existing software because the decision-maki ....Applying forward-backward stochastic differential equations to optimisation. This project intends to develop new ways to solve optimisation problems that are currently difficult to solve because of their complexity and size. In particular, forward–backward stochastic differential equations (FBSDEs) are a new technique that is showing ways to solve problems for which there is yet to be a solution. This project's focus will be on problems that cannot use existing software because the decision-making processes require intensive consideration of all possible outcomes in the modelled environment. In comparison to previous optimisation methods, the FBSDE approach is easier to work with and much more informative. The project's main potential applications are multiplayer games with mean-field interaction and financial markets with partial information.Read moreRead less
A fast comparative method for historical linguistics. Linguists are able to infer ancient histories of languages by a procedure known as the Comparative Method. Its results are used in related studies of human genetic and cultural change. However, the Comparative Method is a manual-only process and thus currently is a bottleneck for the science of unravelling the human past. This project aims to overcome this limitation and significantly accelerate linguistic discovery, by combining recent advan ....A fast comparative method for historical linguistics. Linguists are able to infer ancient histories of languages by a procedure known as the Comparative Method. Its results are used in related studies of human genetic and cultural change. However, the Comparative Method is a manual-only process and thus currently is a bottleneck for the science of unravelling the human past. This project aims to overcome this limitation and significantly accelerate linguistic discovery, by combining recent advances in computational language processing, statistics and cultural-evolutionary modelling. By producing innovative mathematical means for rapidly discovering ancient language relationships, it will enable a breakthrough in our capacity to uncover human linguistic, genetic and cultural heritage worldwide.Read moreRead less