IEEE CEC 2026 Competition on Multimodal Multiobjective Optimization
(with potential cash prize💰)
IEEE CEC 2026 Competition on Multimodal Multiobjective Optimization
(with potential cash prize💰)
Multimodal Multiobjective Optimization (MMO) aims to detect and approximate the entire Pareto Set of a multiobjective optimization problem, even when a portion of it can represent the entire Pareto Front. It is a relatively recent field of research in evolutionary computation that combines the strengths of two well-established areas: multimodal optimization and multiobjective optimization. Like multiobjective optimization, it offers a set of solutions that highlight the best trade-offs among problem objectives. Like multimodal optimization, it provides the decision maker with distinct alternative solutions for a selected trade-off among objectives.
This research field has seen a growing number of new methods, typically formed by integrating a niching strategy into a multiobjective optimizer. However, special attention should be paid to reliable, comprehensive, and pragmatically sound performance assessment and comparison of these methods to guide this emerging field in the right direction. This competition aims to fulfill that goal by providing a fair and thorough experimental setup for evaluating and comparing MMMOO methods.
Competition Outline
Participants should use their methods (new or existing) to optimize the set of black-box MMO test problems of the competition with up to 5 objectives and 30 decision parameters. The results should be submitted as CSV files, which will be used to rank the participating methods using the Friedman test. The code for the test problems and the performance indicator is provided as MATLAB p-files. For each run, a lower output number for the performance indicator means a better performance.
Participants should submit the results of their method on all problems, as instructed in the competition’s experimental setup, in CSV format. Additionally, participants should provide either a reference to a relevant publication or a PDF document explaining their method. The results and the method description will be made publicly available.
Submission Instructions
Participants should submit the required documents, as instructed in the competition technical report. Each submission should include:
The names and affiliations of all participating team members
The result files (1480 .mat files), which are the reported set of solutions by the method for each run.
A PDF document describing the methods used
Participants are also strongly encouraged to publicly share their code. If public sharing is not possible, they should clearly state that their code will be made available privately upon request in the PDF file. These documents will be shared publicly on the competition website after the conference.
For participation, please submit your package to Ali Ahrari (a.ahrari@unsw.edu.au, aliahrari1983@gmail.com) by 1 June 2026.
If you plan to participate, please send us an email (to the email(s) mentioned above) so that we can share the latest update with you.
Download Links
MATLAB code (Test problems + Performance calculator) (Style updated on 9-April-2026) (download)
MATLAB code - Augmented on 15-April-2026: An example was added to show how to optimize a problem with a trivial optimizer (download)
Sample results (download)
Technical Report for the competition description (on Zenodo) (direct download)
Organizers
UNSW Canberra
a.ahrari@unsw.edu.au
RMIT University
xiaodong.li@rmit.edu.au
Michigan State University
kdeb@egr.msu.edu
Competition Outcome (To be announced at the IEEE CEC 2026 conference):
First place:
Second Place:
Third Place: