Evolutionary Computation and Decision Making
Evolutionary Computation and Decision Making
Keynote talk: Prof. Ed Keedwell, University of Exeter, UK
Title: Human-AI Interaction in Multi-Criteria Decision Making for Real-World Applications
Abstract: Multi-objective metaheuristics have a long history of producing excellent solutions for problems with multiple objectives. However, recent work has shown that these solutions can achieve improved feasibility in real-world applications through the incorporation of domain or user expertise. In this talk I will describe the development of automated heuristics and their incorporation in an evolutionary algorithm that improve the feasibility and mathematical optimality of the developed solutions for water distribution network (WDN) optimisation. I will then go on to describe work on a new interactive evolution system that uses visualisation, interaction and machine learning to automate the extraction of human knowledge and embed it in a multi-objective evolutionary algorithm using WDN optimisation as an example.
Accepted papers