Keynote Details

We are pleased to confirm that our keynote talk will be delivered by Prof. Jim Smith from the University of the West of England, Bristol, UK.

Interactive Memetic Algorithms for optimisation, design, and modelling

Interactive Evolutionary Algorithms have been successfully applied in many applications where the measure of ‘quality' is a user’s satisfaction. However, people often get bored, or frustrated by over-simplistic tools, and even with the best intentions, their judgements become less consistent over time.

Key to maintaining engagement is to offer a richer form of human-machine teaming that can: support other paradigms such as one-shot learning; take account of human frailties; and above all offer more transparent human-machine collaboration.

Using a series of examples drawn from different application domains, I will argue that the Memetic paradigm offers an ideal way to support a richer range of interactions. Interactive MAs can be applied to ’straight-forward’ optimisation/design problems, and also to model creation within ‘active machine learning’.

I will also show how incorporating different multi-modal interfaces such as eye-trackers let us create tools that can serve a range of goals: from rapid optimisation to prolonged exploration; from extreme personalisation, to the creation of shared knowledge in a multiple-party human-machine team.

Comments

Post a Comment

Popular posts from this blog

iGECCO – A Workshop on Interactive Evolution at GECCO 2019

Image
Nature-inspired algorithms are powerful problem-solvers where a mathematical formulation of a problem can be developed.  However, for some applications, particularly in the arts and humanities, human subjectivity is required to enable the algorithms to produce useful outputs for human consumption.  Examples of this exist in evolutionary art, music and product design where the challenge for researchers is to develop systems that will make best use of user inputs but reduce human fatigue in interacting with the algorithm.  Furthermore, interactive evolution is beginning to be used in other, more scientific domains where the expression of objectives and constraints in a computable function is non-trivial or impossible.  This workshop will explore the development of nature-inspired techniques where human interaction plays a part in shaping the algorithm, solution development or selecting final solutions that meet human subjective criteria.  We aim to host an exciting and varied event at
Read more

iGECCO 2023 - Call for Papers

  Aim and Scope As nature-inspired methods have evolved, it has become clear that optimising towards a quantified fitness function is not always feasible, particularly where part or all of the evaluation of a candidate solution is inherently subjective. This is particularly the case when applying search algorithms to problems such as the generation of art and music. In other more traditional application areas, optimising to a fitness function might result in a highly optimal solution that is not well suited to implementation in the real world. Incorporating human expertise into the optimisation process can yield useful results in both examples, and accordingly the work on interactive optimisation methods has matured in recent years. This workshop will provide an outlet for interactive optimisation research for the GECCO audience. Topics of interest include (but are not limited to):    * Interactive generation of solutions.  * Interactive evaluation of solutions.  * Psychological aspect
Read more

iGECCO 2020 Announcement

Image
We are excited to announce that the Interactive Methods workshop will be returning to GECCO for a second year running. This year's organising committee is: Matthew Johns (University of Exeter) Nick Ross (University of Exeter) Ed Keedwell (University of Exeter) Herman Mahmoud (University of Exeter) David Walker (University of Plymouth) More information can be found on the GECCO 2020 website here . 
Read more