Exploring the emergence of collective behavior through neuroevolutionHisham Radwan, Ahmed (Avtor) Lebar Bajec, Iztok (Mentor) Demšar, Jure (Komentor) Collective behaviorNeuroevolutionEvolutionThis thesis investigates the emergence of collective behavior in a simu- lated predator-prey ecosystem using neuroevolution, specifically the NEAT algorithm. The study employs a multi-stage evolutionary process, including independent evolution and coevolution of predator and prey agents. Re- sults demonstrate the effectiveness of neuroevolution in evolving adaptive behaviors, with predators improving hunting strategies and prey develop- ing typical patterns of collective behavior, such as swarming. The research contributes to understanding the adaptive significance of collective behavior in predator-prey systems and highlights the potential of neuroevolution in studying emergence of complex behavior in real world.20242024-11-25 08:00:18Magistrsko delo/naloga165149VisID: 37157COBISS_ID: 217243395sl