PPSN 2026 Workshop: Tailoring Optimization Approaches with Domain Knowledge¶
Workshop Description¶
Optimization research has traditionally focused on improving algorithms from a theoretical perspective or through benchmarking studies. While this has led to powerful general-purpose methods, practitioners solving real-world problems rarely face well-defined benchmark settings. Instead, optimization problems arrive in messy forms: objectives may be unclear, constraints incomplete, data uncertain, and evaluation pipelines complex. In practice, success often depends less on selecting the “best” algorithm and more on understanding the application and tailoring the optimization process accordingly.
Currently, there is no systematic methodology for such tailoring. Existing approaches typically move in one of two directions: adapting the problem formulation to match a given algorithm, or modifying the algorithm to fit a specific problem. As a result, knowledge gained from applications remains fragmented, making it difficult to transfer experience across domains and slowing scientific progress. Although classical textbooks provide general advice, they predate modern optimization paradigms such as Bayesian optimization, automated configuration, large-scale benchmarking, and hybrid learning-optimization pipelines.
This workshop aims to develop a shared understanding of how domain knowledge can be systematically incorporated into optimization. We will bring together researchers and practitioners from evolutionary computation, Bayesian optimization, machine learning, and application domains to collect, structure, and generalize existing experience. Rather than presenting isolated success stories, the goal is to identify recurring patterns: how problems are modeled, how representations and constraints are chosen, how uncertainty and evaluation cost shape the algorithmic design, and how optimization fits into real decision-making workflows. The workshop will combine invited perspectives, a panel, contributed talks, and interactive discussions. Participants will collaboratively organize practical insights into a structured and searchable body of knowledge describing tailoring strategies that apply across optimization methods. The workshop seeks to democratize access to expertise and enable both researchers and practitioners to design more effective problem-aware optimization processes.
Ultimately, the outcome will be a step toward a modern “recipe book” for real-world optimization, bridging theory and practice and enabling systematic progress beyond benchmark-driven algorithm development.
Target Audience¶
The workshop targets researchers and practitioners working on optimization for real-world problems, including evolutionary computation, Bayesian optimization, metaheuristics, and automated machine learning. It is particularly relevant for participants interested in moving beyond benchmark settings toward application-driven algorithm design. We welcome both algorithm developers seeking principled ways to incorporate domain knowledge, and application experts facing complex optimization pipelines in practical fields.
Planned Schedule¶
Duration: 2 hours
Format: In-person
| Time Allocation | Activity |
|---|---|
| 5 minutes | Opening Remarks |
| 25 minutes | Invited Talk |
| 45 minutes | 3 Contributed Talks |
| 30 minutes | Panel Discussion |
| 15 minutes | Interactive Discussion & Closing |
Submission Details¶
We solicit submissions in the form of extended abstracts (max 4 pages, references included) describing real-world optimization problems, modeling choices, or tailoring experiences. Contributions may include practical case studies, methodological insights, or reflections on challenges encountered in real-world settings. Please follow the PPSN submission guidelines.
Accepted contributions will be presented as short talks and will feed into the interactive sessions and panel discussions of the workshop.
Please submit a PDF document by email to Elena Raponi (e.raponi@liacs.leidenuniv.nl) with a message header following the format [PPSN26: TAILORING WORKSHOP].
Important Dates¶
- Submission deadline: June 5, 2026
- Notification: June 30, 2026
- Camera-ready: July 31, 2026
- Author's mandatory registration: July 31, 2026
- Workshop: During PPSN 26, August 29 / 30, 2026
Contact details and short CV of the organizers¶
Elena Raponi is an Assistant professor in Bayesian optimization at the Leiden Institute of Advanced Computer Science (LIACS) of Leiden University, in the Natural Computing research cluster. Previously, she held postdoctoral positions at LIACS, the Technical University of Munich (TUM), and Sorbonne Université. She received her PhD in Applied Mathematics from the University of Camerino, Italy, in 2021. Her research focuses on surrogate-based optimization, with a particular emphasis on constrained, high-dimensional Bayesian Optimization in continuous domains, and automated algorithm design and configuration assisted by LLMs. She works on the development of analytical and numerical modeling techniques for optimization in structural mechanics. Her hybrid research profile enables algorithm design inspired by concrete challenges emerging from real-world applications. She co-organized Dagstuhl seminars on AutoML and engineering design, as well as on leveraging domain knowledge in optimization. She also served on the organizing team of the AutoML Conference (2024–2025) and is Program Chair of the upcoming edition.
Affiliation: Leiden University (LIACS)
Email: e.raponi@liacs.leidenuniv.nl
Website: https://www.universiteitleiden.nl/en/staffmembers/elena-raponi#tab-1
Google scholar: https://scholar.google.com/citations?user=puWIVC4AAAAJ&hl=e
Vanessa Volz is currently a tenure track researcher in the Evolutionary Intelligence (EI) group at Centrum Wiskunde & Informatica (CWI), Amsterdam, The Netherlands. She received her PhD in 2019 from TU Dortmund University, Germany, for her work on surrogate-assisted evolutionary algorithms applied to game optimisation. She holds B.Sc. degrees in Information Systems and in Computer Science from WWU Münster, Germany. She received an M.Sc. with distinction in Advanced Computing: Machine Learning, Data Mining and High Performance Computing from University of Bristol, UK, in 2014. Her current research focus is on employing surrogate-assisted evolutionary algorithms to obtain balance and robustness in systems with interacting human and artificial agents, especially in the context of games.
Affiliation: Centrum Wiskunde & Informatica (CWI)
Email: Vanessa.Volz@cwi.nl
Website: https://www.cwi.nl/en/people/vanessa-volz/
Google scholar: https://scholar.google.com/citations?user=S_l5rNYAAAAJ&hl=de
Carola Doerr, formerly Winzen, is a CNRS research director at Sorbonne Université in Paris, France and a scientific advisor for CNRS Informatics. Carola's main research activities are in the analysis of black-box optimization algorithms, both by mathematical and by empirical means. Carola is associate editor of IEEE Transactions on Evolutionary Computation, ACM Transactions on Evolutionary Learning and Optimization (TELO), and the Evolutionary Computation journal. She was program chair for AutoML 2025, for the BBSR track at GECCO 2025 and 2024, the GECH track at GECCO 2023, for PPSN 2020, FOGA 2019, and for the theory tracks of GECCO 2015 and 2017. She has organized Dagstuhl seminars and Lorentz Center workshops. Together with Pascal Kerschke, Carola leads the 'Algorithm selection and configuration' working group of COST action CA22137. Carola's works have been distinguished by several awards, among them the CNRS bronze medal, the Otto Hahn Medal of the Max Planck Society, and best paper awards at GECCO, CEC, and EvoApplications.
Affiliation: CNRS, Sorbonne Université (LIP6)
Email: Carola.Doerr@lip6.fr
Website: https://doerr.perso.lip6.fr/
Google scholar: https://scholar.google.com/citations?user=CU-V1sEAAAAJ&hl=it
Joshua Knowles is a scientific advisor with SLB (Schlumberger) in Cambridge, UK where he works on scalable energy solutions and optimization. He is also an honorary professor at the Alliance Manchester Business School, University of Manchester, and honorary senior fellow (formerly full professor) at the School of Computer Science, University of Birmingham. Josh has a wide range of industrial and interdisciplinary experience applying evolutionary computing and ML methods in areas such as analytical biochemistry, astrophysics, telecoms, food science, systems biology, and electricity trading. He was a keynote speaker at FOGA2025, general co-chair of EMO2025, a tutorial speaker at GECCO2024, and has run workshops at GECCO, CEC, and PPSN since the late 1990s. He was co-organizer of a seminar on tailoring with domain knowledge at Dagstuhl in February 2026, his fourth Dagstuhl Seminar. Josh is also an IEEE Fellow, cited for “contributions to multiobjective optimization”, and has twice won the IEEE CIS Outstanding Paper Award of the Transactions on Evolutionary Computation, as well as the ACM SIGEVO Impact Award.
Affiliation: SLB (slb.com)
Email: knowles.joshua@gmail.com
Website: https://www.researchgate.net/profile/Joshua-Knowles?ev=hdr_xprf\
Google scholar: https://scholar.google.com/citations?hl=en&user=nltQkfgAAAAJ