Thema der Dissertation:
Topics in particle-based simulation: Multiscale Models and optimization methods Thema der Disputation:
Multiscale Search in the Bees Algorithm
Topics in particle-based simulation: Multiscale Models and optimization methods Thema der Disputation:
Multiscale Search in the Bees Algorithm
Abstract: The Bees Algorithm (BA) is a population-based optimization method inspired by the foraging behavior of honey bees. Originally developed for continuous optimization, it has since been applied to a wide range of problems, including combinatorial optimization and machine learning. In comparison to other swarm-inspired algorithms, the exploitation process is governed not only by fitness, but by an explicit neighbourhood that is progressively shrunk and re-centered around improved solutions. New candidate solutions are then sampled at random within this neighbourhood, which effectively controls the scale at which the search space is explored. In this talk, I will focus on how this multiscale search influences convergence, exploration, and stagnation. I will further illustrate how these mechanisms affect performance on different types of optimization landscapes and how problem-specific modifications can improve its effectiveness.
Zeit & Ort
08.05.2026 | 12:00
Hörsaal (Raum Nr. 2005)
(Zuse-Institut Berlin, Takustr. 7, 14195 Berlin)