Research Themes
The Centre for Augmented Reasoning (CAR) will enable the University of Adelaide to grow its advanced machine learning capability with grants to support the delivery of a world-class research program focusing on natural language processing, computer vision, deep learning, and machine reasoning.
The centre’s ambitious research plan identifies four main research themes that will chart the course for the next generation of Australian AI leadership.
Theme 1: Next Generation Machine Learning
We will improve the underlying tools to enable more functional AI to deliver better solutions by improving the efficiency of learning, enabling learning at scale, building next generation causality capability and applying quantum technologies to reasoning.
Theme 2: Interactive Machine Learning
Our researchers will develop new models for reasoning and storing information that can be called upon by machines to solve new challenges. This includes developing advanced vision-and-language systems, and building machines that better understand humans and that can learn by interacting with us and the environment.
We will explore new areas of vision-and-language research, build novel algorithms, and collaborate with world-renowned musicians and artists to develop new ideas and creations at the frontiers of music, art and AI.
Theme 3: Knowledge, Representation and Generalisation
We will build the technology to enable machines that are more capable of learning on their own and can explain their reasoning. We will develop more adaptable learning systems that can be applied to new examples with minimal retraining. This includes developing novel semantic representations for natural languages and methods that can generalise across scenarios and support complex reasoning.
Theme 4: Machine Learning Driven Science Discovery
Using machine learning to support discoveries in biological and chemical sciences, our researchers will convert our research capability into solutions for the intractable challenges faced by humans now and in the future. This includes predicting treatment outcomes for patients by testing and implementing machine learning algorithms, and leading AI-driven discovery for energy materials.