Beyond a Performance Scalar at EvoStar 2027
We are excited to announce the special session Beyond a Performance Scalar: Evaluation Science for Evolutionary Computation at EvoStar 2027 (EvoApps).
Official website: https://www.evostar.org/2027/evoapps/beyond-a-performance-scalar-evaluation-science-for-evolutionary-computation/
Evaluation Science for Evolutionary Computation
This special session invites original research on evaluation methods that go beyond single summary numbers for evolutionary and nature-inspired algorithms, such as final best fitness, mean or median over runs, or ranking tables. It covers three levels of evidence:
- Theory and mechanisms: convergence, runtime, stability, population dynamics, exploration versus exploitation, diversity, and component contributions.
- Benchmarks: property- and distribution-aware evaluation across landscapes, budgets, constraints, noise, dynamic changes, and fidelity levels.
- Applications: evaluation that combines solution quality with feasibility, reliability, uncertainty, interpretability, cost, resource use, and decision relevance.
Papers whose main contribution is a new optimizer judged only by final-fitness tables are outside the scope of this session. We seek methodological work backed by empirical or theoretical validation.
Topics of Interest
- Anytime, time-to-target, and convergence-profile analysis
- Behavioural, trajectory, and diversity measures
- Ablation, sensitivity, and mechanism attribution
- Uncertainty, effect sizes, statistical comparison, and ranking stability
- Distributional and multi-criteria alternatives to scalar aggregation
- Explainable performance analysis
- Robustness and generalization
- Evaluation of multimodal, constrained, dynamic, noisy, and multi-fidelity problems
- Cost- and decision-aware evaluation in real-world applications
- Evaluation of evolutionary machine learning, AutoML, automated algorithm design, and LLM-assisted evolutionary computation
- Frameworks linking theory, benchmarks, and practical outcomes
Key Dates
- Paper Submission Deadline: 1 November 2026
- Notification of Acceptance: 10 January 2027
- Camera-ready Submission: 24 January 2027
- EvoStar 2027 Conference: 31 March - 2 April 2027, Mainz, Germany
Organisers
- Rohit Salgotra, AGH University of Krakow, Poland
- Tome Eftimov, Jožef Stefan Institute, Ljubljana, Slovenia
- Niki van Stein, Leiden University, the Netherlands
- Anja Jankovic, RWTH Aachen University, Germany
- Eva Tuba, Trinity University, San Antonio, TX, USA, and Jožef Stefan Institute, Ljubljana, Slovenia
Acknowledgements
This event is made possible by the IEEE CIS Task Force on Generative AI for Evolutionary Computation (GenAI4EC).
Organizers:
