Author: oneillm

Today the UCD NCRA celebrates it’s 20th anniversary. It was founded on 1st January 2006 by Prof Anthony Brabazon and Prof Michael O’Neill when Michael joined the faculty at UCD. There will be a number of activities to mark this anniversary over the course of 2026. First up will be Prof Franz Rothlauf who will visit our group for a number of weeks from mid-January to the end of February.

We would like to take this occasion to acknowledge and thank all the current and past members of the UCD NCRA with whom we jointly celebrate this occasion. Without our PhD students, Postdocs and our numerous academic and industry collaborators we would not be celebrating so much success to date.

For posterity below is an image from the archive (February 2006) when we marked the foundation of the group and simultaneously launched the publication of our book “Biologically Inspired Algorithms for Financial Modelling”. We were and still are deeply grateful for the support we received in forming the group. In the photo we are joined by three of these key people, Prof Hugh Brady (then UCD President – second from the right), Prof Barry Smyth (then Head of School of Computer Science – center/third from the right), and Prof Tom Begley (then Dean of Business – first from the left).

Our group was well represented at GECCO in Malaga this July. Miguel organised the Summer School, and we had two papers published (double-blind peer review) by our PhD students Jiajun and Zhiang both reporting on research undertaken as part of the Research Ireland funded RegFrAIMs project. Jiajun presented “On the use of Hinge Loss as a Surrogate Fitness Function with Grammatical Evolution for Parkinson’s Disease Classification“. And, Zhiang presented “Investigating Combined Algorithm Selection and Hyperparameter Optimization for Fairness“.

Dr Mark Connor organised and Chaired the inaugural UCD Smurfit Sports Analytics Conference, which was held in the UCD Smurfit School on 9th July 2025. An exciting programme included two keynotes, two panels (AI in Sport, and Digital Transformation in Sport), and set of lightening research talks. The keynote speakers were Dr Kevin de Keizer, (Manchester United Women’s Academy), and Ruth Montgomery (Irish Rugby Women’s National Team).

Our visiting PhD researcher, Romain Claret successfully published and presented his latest research on “Investigating Hyperparameter Optimization and Transferability for ES-HyperNEAT: A TPE Approach” at the recent ACM GECCO conference held in Melbourne Australia (14-18 July 2024).

A new part-time Professional Diploma in Artificial Intelligence & Business Analytics is accepting applications to commence at the end of April 2024, and features faculty from the UCD Natural Computing Research & Applications group. This is a fully accredited (NFQ Level 9 / 30 ECTS) PDip with six modules. Each module runs over two days on site in the Smurfit School, with modules running in April, June, Sep, Oct, Nov, and Jan.

Targeted at business executives, the diploma will provide participants with the knowledge, tools, and insights to demystify AI and business analytics and explore their application in their organisation. Real-world examples will be presented to demonstrate the capabilities, limitations, and opportunities provided by these technologies. Participants will discover how to leverage these to create a competitive edge for their business.

Prof’s O’Neill & Brabazon have published a chapter on Evolutionary Machine Learning in Finance (Chapter 24) as part of the new Handbook of Evolutionary Machine Learning edited by Wolfgang Banzhaf, Penousal Machado, Mengjie Zhang

Dr Mark Connor was interviewed on the future of AI for Authority Magazine

Congratulations to Dr Mark Connor and Dr Nam Le Hai who graduated with PhD’s from University College Dublin and the UCD Natural Computing Research & Applications Group on 5 September 2023. Dr Mark Connor’s PhD thesis title is Optimisation and Adaptation of Athlete Training Load with Evolutionary Computation. Dr Nam Le Hai’s thesis is The Interplay between Evolution and Learning in Dynamic Environments: Computational Metaphor of Adaptive Algorithms

Following a recent podcast interview and conversations with colleagues outside of the field of Artificial Intelligence, Michael and Mark decided to collate a list of limitations of large language model technology, which some might find useful. These are captured in their new article Amplifying Limitations, Harms and Risks of Large Language Models published on arXiv

Congratulations to Dr Miguel Nicolau who is a Secondary Proposer for a newly awarded  COST Action (CA22137) “ROAR-NET – Randomised Optimisation Algorithms Research Network“