Collaborative Machine Learning for Research
Abstract
I will bring together two traditionally separate lines of research: (i) probabilistic machine learning, in which we combine flexible learning from multiple sources of data with prior knowledge in the form of simulators, and (ii) user interaction, starting from interactive intent modelling in information retrieval and extending to collaborative AI. This combination enables developing new kinds of tools for research tasks, including AI-assisted design tools for design-build-test-learn cycles. When R&D processes are formulated as virtual simulation-based laboratories, the tools can be applied across fields from experimental sciences to engineering design and medicine, and further to humanities and social sciences. The university is now building significant capacity in this field with the launch of the Centre for AI Fundamentals, AI-FUN, with research groups focusing on developing machine learning principles and seeking to do that in collaboration with other fields.
