Scalable Neuroevolution of Ensemble Learners
Talk, University of Bremen, Group of Computer Architecture, Bremen, Germany
Together with Marcel Merten, I presented our work Scalable Neuroevolution of Ensemble Learners as a Group of Computer Architecture Talk at University of Bremen. In the context of evolutionary optimization, it addresses one of its main issues: lack of scalability. To this end – based on ALF – a framework was developed to heavily orchestrate genetic recombination for neural network combinations to effectively find suitable solutions. Moreover, ensemble learning methods restrict the search space. In a divide-and-conquer manner, the framework allows handling of large-scale machine learning problems.