Multimodal optimization is a growing field of research, motivated by the importance of identifying diverse and alternative solutions -- in addition to the single best one -- for real-world problems. Since mathematical optimization problems are typically simplified models of actual problems, they may -- inadvertently or for simplification purposes -- fail to capture all aspects of the real-world scenarios. The availability of diverse (near-)optimal solutions provides decision-makers with the opportunity to manually select the most suitable solution for the actual problem -- not just the modeled one -- based on their expertise and experience.  

This competition employs an updated test suite that includes fully scalable and tunable test problems that simulate diverse challenges associated with multimodal optimization. When compared to the similar competitions organized at GECCO in 2024 and 2025, the updated test problems can also simulate diverse types of linkage among decision variables -- a feature that did not exist in the previously held competitions -- such that the test problems include completely separable, partially separable, and fully correlated ones.  

A user-friendly platform for the test problems will be provided in MATLAB and Python, which can easily be integrated with almost any existing multimodal optimization method.  

Download Links

MATLAB Code 

Python Code

Sample Result Files

Competition Description

Organizers

Competition Outcome:

First place:  

Second place: 

Third place: