Population-based evolutionary search for joint hyperparameter and architecture optimization in brain-computer interface

Dong-Hee Shin, Deok-Joong Lee, Ji-Wung Han, Young-Han Son, Tae-Eui Kam

Abstract: Highlights•We propose to optimize both hyperparameter values and neural network architectures.•Population-based evolutionary search enables an adaptive optimization in parallel.•Two-stage optimization process effectively handles the complex joint search space.•Ensemble method further improves the robustness and generalization capabilities.•Our method outperforms all competing methods across a range of backbone models.