Evolutionary Computation and Decision Making

Evolutionary Computation and Decision Making

Keynote talk

Keynote talk: Michael Kirley, School of Computing and Information Systems, University of Melbourne.

Title: Evolutionary Computation and uncertainty-based Optimisation for Multi-objective Decision Making

Abstract: In this presentation, I will discuss how metaheuristics, Bayesian optimisation and multi-fidelity simulations (mechanistic and agent-based models) can be used to predict and optimise processing pipelines. A number or uses cases, spanning biopharmaceuticals, rheology, urban freight systems and agri-business industrial applications will provide the scaffolding for this presentation. Two themes will be examined in detail: (1) optimisation under uncertainty, including multi-objective problems and optimisation in dynamically changing environments, and (2) analysing experimental data with advanced statistical, optimisation and active learning techniques. Importantly, the strength and limitations of specific methodologies will be discussed, as well as the need to account for explainability in optimisation and decision-making

Bio: Michael Kirley is a Professor in the School of Computing and Information Systems and a Director of the Melbourne Centre for Data Science at the University of Melbourne. His primary area of expertise is positioned at the intersection of AI, machine learning and data science, with a strong emphasis on evolutionary computation (EC). Michael’s main scientific contributions include the design and analysis of evolutionary algorithms with work focused on solving large-scale problems, with multiple competing objectives and uncertainties, which require decisions on multiple time-scales. His research has also contributed to a more nuanced understanding of how AI tools interact and possibly deceive. Michael has attracted $AUD 13M in grant funding, including five ARC Discovery Projects and multiple Australian Government Innovation grants. He is currently a chief investigator in the ARC Industrial Transformation Research Hub for Digital Bioprocess Development and the ARC Industrial Transformation Training Centre in Optimisation Technologies, Integrated Methodologies, and Applications (OPTIMA). Michael has previously been an Associate Editor for the IEEE Transactions on Evolutionary Computation and has served in numerous roles for leading EC and AI conferences.