Ensemble deep random vector functional link for self-supervised direction-of-arrival estimation

Jiawen He, Xiaolei Li, Peishun Liu, Liang Wang, Hao Zhou, Jinyu Wang, Ruichun Tang

Abstract: Highlights•We design a signal model-embedded loss function, which is a closed-form solution.•We introduce an L1 auxiliary loss term to obtain the signal sparse solution.•The proposed models are instantiated and have excellent generalization ability.•We compare with the RVFL variants, the results manifested competitive performance.