Evolution of Artificial Life
I carried out my BSc thesis, Evolution of Artificial Life: Investigating Optimisation in Gene Space, under the supervision of Dave Clements in the Astrophysics Group at Imperial College London. My partner Thushanan and I developed a numerical agent-based Python simulation to investigate the optimisation problem of evolution. The project was awarded a distinction.
The viva slides can be found here and the report here.
Abstract:
This project investigates the optimisation problem of evolution by using a numerical agent-based simulation. The results provide valuable insights into the dynamics of this process in nature. After validation using the Lotka-Volterra model, we examined reproductive strategies and vision in a single-species system. Our findings confirmed hypotheses about the effect of evolutionary pressures on reproduction rates and optimal gene values for food distribution. We also observed the positive impact of mutations on survival rates, as well as coevolutionary dynamics in predation, namely through speed and colour. These results aim to demonstrate the adequacy of the evolutionary algorithm to meet the criteria enforced by natural selection and the role of mutations and pressures in shaping the genetics of populations over time. This project provides a possible starting point for further investigation into computational modelling of evolution.