Short-Term Scientific Mission

Development of a General-Purpose Evolutionary Algorithm Based on the ROAR-NET API

Main theme: Problem Modelling and User Experience
Grantee: Nathan Amoussou, Université Cöte d'Azur, Sophia Antipolis, France
Host: Carlos Fonseca, University of Coimbra, Coimbra, Portugal

Start date: 2026-09-14
End date: 2026-09-30
Awarded: 2026-07-30

Description

The main goal of this STSM is to develop a general-purpose evolutionary algorithm based on the ROAR-NET API. A combinatorial optimisation problem for which evolutionary methods are known to be competitive will be selected, and a corresponding ROAR-NET API problem model will be implemented, followed by a mutation-based evolutionary algorithm and distance-based mechanisms capable of supporting the implementation of geometric recombination operators on the algorithm side. The resulting model and algorithm implementations will be validated and compared with a credible reference method from the literature with respect to solution quality, computational effort and evaluation cost. The STSM aims to provide reusable software, benchmark results, and practical feedback on the ROAR-NET API from a user’s perspective.