Researchers from CWI won, in teams with others, three awards for AI in healthcare at the ACM Genetic and Evolutionary Computation COnference (GECCO) 2021. These awards were: the Best Paper Award in the Genetic Algorithms track for a team of CWI, TU Delft and LUMC, the SIGEVO Dissertation Award 2021 for CWI’s Marco Virgolin and a Silver HUMIES award for CWI and Amsterdam UMC. GECCO is the premier conference in the field of evolutionary algorithms. The 2021 edition was held online in July.
Best Paper Award
Two researchers of CWI’s Life Sciences and Health (LSH) group - Arkadiy Dushatskiy and Peter A.N. Bosman (CWI/TU Delft) - together with Tanja Alderliesten (LUMC) received the Best Paper Award in the Genetic Algorithms track. They were awarded for their paper ‘A Novel Surrogate-assisted Evolutionary Algorithm Applied to Partition-based Ensemble Learning’.
Computationally expensive optimization occurs in various domains, like neural architecture search and simulation-based optimization. To improve this, the researchers propose a novel surrogate-assisted Evolutionary Algorithm (EA), based on the long-running Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) research line in their group. The new EA shows better performance compared to simple search baselines, plain GOMEA, and two discrete Bayesian optimization algorithms. Currently it is being studied within the context of deep learning.
SIGEVO Dissertation Award for Marco Virgolin
The yearly dissertation award of the ACM Special Interest Group on genetic and EVOlutionary computation (SIGEVO) recognizes excellent thesis research by doctoral candidates in the field of evolutionary algorithms. A committee of renowned researchers reviewed the nominated dissertations for technical depth and significance of the research contribution, potential impact on the field, and quality of the presentation. CWI’s Marco Virgolin won the SIGEVO Dissertation Award 2021.
As reported by the committee: "Marco’s dissertation introduces significant innovations in model-based genetic programming (GP) enabling the efficient discovery of compact and accurate solutions, which is key for supporting explainable Artificial Intelligence. In collaboration with medical researchers, the proposed innovations are applied successfully to improve historical 3D-dose reconstruction in radiation oncology which is used to design more effective treatments for cancer patients. This work is a stellar example of the positive real-world impact of Evolutionary Computation."
Silver HUMIES award for research from a project at CWI and Amsterdam UMC
A team of seven international institutes, led by Amsterdam UMC – location AMC - and CWI, has been awarded the Silver Award in the HUMIES competition – the competition on results obtained by evolutionary computation that are on par with, or better than, those obtained by humans on recognized problems or challenges in a field. From Amsterdam UMC former PhD student Ziyuan Wang and project leader Tanja Alderliesten were predominantly involved - both also CWI guest researchers – and from CWI especially former PhD student Marco Virgolin and his supervisor Peter A.N. Bosman
The award was granted for the work done on a Stichting KiKa funded project between Amsterdam UMC and CWI on improving, by means of evolutionary machine learning, the state-of-the-art in radiation dose reconstruction, an important problem that is key to better understanding the long-term effects of radiation treatment for cancer. Researchers at CWI invented a novel evolutionary machine learning algorithm (GP-GOMEA), and, jointly with researchers at AMC, designed novel automatic pipelines that leveraged GP-GOMEA to automate the typically mostly manual dose reconstruction process and improve its accuracy. The collaboration with various international institutes made it possible to validate the approach, showing its capability to perform dose reconstruction on par or better than humans, while dramatically reducing the required manual labor.
See: the presentation video by Marco Virgolin.
About GECCO
The ACM Genetic and Evolutionary Computation Conference (GECCO) is the premiere conference in the field of evolutionary algorithms and yearly presents the latest high-quality results in genetic and evolutionary computation since 1999. Topics include: genetic algorithms, genetic programming, ant colony optimization and swarm intelligence, complex systems (artificial life, robotics, evolvable hardware, generative and developmental systems, artificial immune systems), digital entertainment technologies and arts, evolutionary combinatorial optimization and metaheuristics, evolutionary machine learning, evolutionary multiobjective optimization, evolutionary numerical optimization, real world applications, search-based software engineering, theory and more.
More information
- Peter Bosman's contact details at CWI
- CWI's Life Sciences and Health research group
- GECCO 2021
- All presentations for the HUMIES Awards (Presentation by Marco Virgolin)
- In 2019 Peter Bosman and others also won a Silver HUMIES Award
Picture: Peter Bosman (CWI and TU Delft) watching the three online GECCO 2021 award ceremonies for his team and other colleagues. Peter Bosman comments: "Watching it unfold from home is not the same, but still: humbled & proud of all involved (especially Arkadiy Dushatskiy, Marco Virgolin, Ziyuan Wang and Tanja Alderliesten) to see 3 awards at the premiere conference on evolutionary computation (GECCO) come our way this year!"
(Own picture.)
