Strategic decisions for competitive and complex games
Project Leader: Tansu Alpcan
Collaborators: Alex Kalloniatis (DST Group)
Primary Contact: Tansu Alpcan (firstname.lastname@example.org)
Keywords: complex systems; game theory; network science; optimisation
Disciplines: Electrical & Electronic Engineering
Decisions of competing teams and organisations can be modelled using continuous Perception-Action cycles and synchronisation on networks. The Boyd “Observe, Orient, Decide, and Act”, or OODA, loop along with Kuramoto models are widely known respective examples that are applicable to competitive economic behaviour on markets, defence, and competitive team sports. The resulting complex dynamical decision systems are naturally investigated under the framework of Game theory, which studies multi-person decisions in both cooperative and competitive processes.
The goal of this project is to develop novel solutions and insights to real-world strategic and adversarial games, where players in multiple teams have coupled decision cycles leading to complex dynamical behaviour. The model-based approach of the project nicely complements popular model-free reinforcement learning methods and will make use of dynamical system, optimisation, and game theories in addition to modern computational tools from machine and deep learning in order to address fundamental research questions.
The interested student should have a solid science or engineering background, with an interest in complex systems, optimisation, and game theory. The student should have an Australian citizenship. In partnership with Defence Science and Technology (DST) Group, the project offers a top-up scholarship, and possible future internship and employment opportunities.