Australian Research Council-Development of Early Career Researchers (ARC-DECRA), 2021
This project developed new tools and methods to improve how computers learn from data, especially when conditions change over time. It produced open-source software, including GenForge for evolving complex mathematical models, GNBG for generating customizable test problems, EDOLAB for teaching and testing algorithms in changing environments, and SPFP for splitting complex data into useful parts. These advances help researchers and engineers build smarter, more reliable systems for tasks like prediction and decision-making, while also supporting education and international collaboration in artificial intelligence. The award and its impact were also highlighted in the University of Technology Sydney (UTS) media release announcing this DECRA: “Big data solutions are in our genes.”
Semantic Preserving Feature Partitioning (SPFP) is an algorithm developed in Python that splits a dataset into equally informative feature subsets, enabling multi-view learning. This approach preserves feature meaning while allowing diverse models to be trained from different views, improving robustness and generalization in genetic programming.
GenForge is a powerful Python package developed in Python and designed to push the boundaries of Genetic Programming (GP) through innovative techniques, such as multi-population approaches and solution forging. Whether you're an academic, researcher, or industry professional, GenForge provides a robust framework for developing, analyzing, and deploying GP-based solutions.
GNBG is a MATLAB-based tool for creating customizable benchmark problems in continuous numerical optimization. It allows precise control over problem characteristics such as dimensionality, modality, separability, and landscape complexity, enabling researchers to design test suites that closely match real-world optimization challenges.
EDOLAB is a MATLAB-based platform for education and experimentation in dynamic optimization. It enables the design, testing, and comparison of algorithms in changing problem landscapes, supporting research on adaptability and robustness in evolutionary computation.
DDG (Dynamic Dataset Generator) is a MATLAB/Python toolset for creating benchmark datasets tailored to dynamic clustering research. It synthesizes evolving mixtures of Gaussian clusters with heterogeneous, controllable changes, such as shifts in location, variations in scale, rotations, and even changes in the number of clusters, so you can rigorously stress-test algorithms for tracking, adaptation, and robustness in non-stationary settings.
Date: July 14–18, 2024
Location: Melbourne Convention and Exhibition Centre, Melbourne, Australia
Co-chaired by Professor Amir H. Gandomi, the Student Workshop at the Genetic and Evolutionary Computation Conference (GECCO) 2024 provided a dedicated platform for undergraduate and postgraduate students to present their research in genetic and evolutionary computation. This multi-session event aimed to support students in developing their scientific work, foster collaboration, and facilitate their integration into the research community. Covering a wide range of GECCO tracks—from genetic programming and evolutionary machine learning to swarm intelligence and real-world applications—the workshop encouraged contributions from emerging researchers as main authors and presenters.
Date: July 11, 2024
Location: UTS CB02, Level 12, Room 225, University of Technology Sydney, Ultimo, NSW, Australia
Workshop on Genetic Programming for Data Analysis (GPDA 2024) Chaired by Professor Amir H. Gandomi, GPDA 2024 was held on July 11, 2024, at the University of Technology Sydney (UTS) as an internal, invitation-only event. The workshop brought together leading international researchers and UTS scholars to discuss cutting-edge developments in genetic programming, with sessions featuring keynote lectures, invited talks, and contributed presentations. Distinguished speakers included Giovanni Squillero (Politecnico di Torino, Italy), Alberto Tonda (INRAE, France), Ting Hu (Queen’s University, Canada), Roman Kalkreuth (RWTH Aachen University, Germany), Szymon Lukasik (NASK – AGH University of Science and Technology, Poland), Erik Hemberg (MIT, USA), and Fabricio Olivetti de França (Universidade Federal do ABC, Brazil). Topics ranged from Turing-complete GP and system identification to multi-population ensembles, Cartesian GP, neural architecture search, and the integration of large language models into GP. The event concluded with a panel discussion on future research directions, followed by a networking banquet.
“GPDA 2024 was a highly productive and timely workshop. It created a very welcoming and stimulating environment that fostered international collaboration and gave me the opportunity to exchange ideas with leading researchers across the globe. The discussions generated new research directions that will strengthen the field and advance the role of genetic programming in data analysis. At this critical juncture for AI development, initiatives like GPDA are essential for that innovative, interpretable, and human-centered approaches are brought to the forefront. I strongly support the continuation of this workshop series, as it will play a key role in shaping the global research agenda, connecting communities, and inspiring both established and early-career researchers.”
“It was a great opportunity to network and exchange ideas with others researchers. The workshop format was excellent, it allowed in depth discussion and feedback of the presentations, as well as invaluable scientific discussions in the breaks. The size enabled me to have in-depth conversations with most of the participants. This is very important and sometimes difficult to achieve in larger conferences. All this made the workshop an excellent academic research and networking experience that is essential for enabling high impact research and collaborations. For example it enabled me to discuss with new colleagues exciting initiatives for workshops and projects regarding generic programming and symbolic regression that we included in a workshop proposal for the major GECCO conference in 2025.”
“The GDPA workshop proved to be an exceptional event, providing an invaluable platform for sharing and exchanging ideas within the Genetic Programming community. It offered me a great opportunity to present my own research and receive insightful comments from other experts in the field. Equally important, I was able to learn about the most recent advancements and innovative approaches that are shaping the future of this domain. The variety and depth of the talks confirmed the relevance and timeliness of the GDPA initiative.I am convinced that establishing GDPA as a recurring series of events will have a highly stimulating effect on the advancement of Genetic Programming and its related areas. By bringing together researchers, practitioners, and students from around the world, the workshop fosters cross-fertilization of ideas and encourages collaboration that transcends institutional and geographical boundaries. It creates a unique momentum that is likely to accelerate both theoretical progress and practical applications. The friendly environment, excellent facilities, and welcoming atmosphere were essential to the workshop’s success. The presence of prestigious guests and leading experts – a direct result of the tireless involvement and commitment of the organizer, Prof. Gandomi – greatly enriched the discussions and broadened perspectives. The professionalism of the organization, combined with the warm and open spirit of the event, made participation in GDPA both intellectually rewarding and personally enjoyable. I look forward to seeing this workshop grow and continue to inspire our community in the years to come.”
“I would like to give a brief statement about the 1st Workshop on Genetic Programming for Data Analysis (GPDA 2024) that took place at University of Technology Sydney organized by prof. Amir Gandomi. The event gathered a fantastic selection of speakers and participants that engaged in fruitful discussions revolving around different topics of Genetic Programming and their applications. As the outcome of the event, from my part it is already fostering the collaboration with different groups and researchers presented at the workshop, promoting advances in the field.”
I participated as an invited speaker to the Workshop on Genetic Programming for Data Analysis (GPDA), organized by Prof. Ganodmi at the University of Technology Sydney in 2024. In my opinion, the event was extremely successful, bringing together selected experts of the Genetic Programming community in a format which fostered a fruitful exchange of ideas and discussions, which is often very limited in larger conferences. During the event I had the opportunity for a long exchange of ideas with Prof. Olivetti from Brazil and Dr. Kalkreuth from Germany, which eventually led to a collaboration. Workshops like GPDA fit a much-needed niche in the domain, and it would be great for GPDA to become a regular event.”
“About one year ago, I had the honor of being the first keynote speaker at the inaugural Workshop on Genetic Programming for Data Analysis (GPDA 2024), organized by Prof. Amir H. Gandomi at the University of Technology Sydney and supported by the Australian Research Council ... the event was exceptionally well-structured, with a diverse program that balanced fundamental advances in genetic programming with real-world applications. What I found most valuable was the depth of interaction the format enabled, fostering genuine dialogue across different research groups and career stages. Prof. Gandomi’s leadership in organizing GPDA created an environment that encouraged openness, collaboration, and forward-looking discussions on the future of genetic programming in data analysis. I am confident that this workshop series will continue to play a pivotal role in shaping the global research agenda and strengthening the GP community.”
Date: March 3, 2025
Location: UTS CB02, Level 12, Room 225, University of Technology Sydney, Ultimo, NSW, Australia
Chaired by Professor Amir H. Gandomi, the second GPDA workshop was held on March 3, 2025, at the University of Technology Sydney (UTS), bringing together researchers to discuss the latest developments in genetic programming for data analysis. The event featured a keynote lecture by Kalyanmoy Deb (Michigan State University, USA) and invited talks by Adriana-Simona Mihăiță, and AIIA Lab's scholars. Sessions covered a range of topics in evolutionary computation, optimization, and applied genetic programming, with opportunities for networking and collaboration throughout the day.
Learn MoreDate: June 30, 2024
Location: Online
Co-organized by Professor Amir H. Gandomi (UTS, Australia), Professor Kalyanmoy Deb (Michigan State University, USA), Dr. Danial Yazdani (UTS, Australia), Dr. Rohit Salgotra (AGH University of Science and Technology, Poland), and Dr. Mohammad Nabi Omidvar (University of Leeds, UK), this GECCO 2024 competition challenged participants to evaluate their global optimization algorithms on a curated set of 24 problem instances generated by the Generalized Numerical Benchmark Generator (GNBG). Spanning unimodal, single-component multimodal, and multi-component multimodal problem types, the test suite incorporated diverse complexities such as ruggedness, asymmetry, conditioning, and deceptiveness. The competition emphasized not only finding optimal solutions but also understanding algorithmic behavior in navigating complex, deceptive landscapes, providing a rigorous benchmark for innovation in numerical global optimization.
Learn MoreDate: January 15, 2025
Location: Online
Organized by Dr. Rohit Salgotra (AGH University of Science and Technology, Poland & UTS, Australia), Professor Amir H. Gandomi (UTS, Australia & Óbuda University, Hungary), and Professor Kalyanmoy Deb (Michigan State University, USA), this competition challenges researchers to evaluate their global optimization algorithms on 24 problem instances generated by the Generalized Numerical Benchmark Generator II (GNBG-II). The suite spans unimodal, single-component multimodal, and multi-component multimodal problems, incorporating diverse complexities such as ruggedness, asymmetry, conditioning, and deceptiveness. Participants aim not only to find optimal solutions but to gain insights into algorithmic behavior across complex and deceptive search landscapes, offering a rigorous benchmark for innovation in continuous numerical optimization.
Learn MoreDate: May 19, 2025 (Submission Deadline)
Location: Online
Organized by Dr. Danial Yazdani (WINC, Australia), Professor Changhe Li (Anhui University of Science & Technology, China), Dr. Mai Peng (China University of Geosciences, China), Dr. Guoyu Chen (Anhui University of Science & Technology, China), Dr. Michalis Mavrovouniotis (Cyprus University of Technology, Cyprus), Dr. Wenjian Luo (Harbin Institute of Technology & Peng Cheng Laboratory, China), Professor Shengxiang Yang (De Montfort University, UK), and Professor Amir H. Gandomi (UTS, Australia & Óbuda University, Hungary), this competition focuses on evolutionary dynamic optimization methods (EDOs) for unconstrained single-objective continuous dynamic optimization problems. Using the Generalized Moving Peaks Benchmark (GMPB), participants tackle 12 benchmark instances with varying complexity—ranging from unimodal to highly multimodal landscapes, symmetric to highly asymmetric, and smooth to highly irregular search spaces. The aim is not only to find high-quality solutions but also to adapt effectively to dynamic environments. Performance is assessed using the offline error metric across 31 independent runs per instance, offering a rigorous platform for fair comparison of EDO algorithms.
Learn More