Papers propose robust RVFL variants for class-imbalanced data

Three arXiv papers propose robust Random Vector Functional Link (RVFL) network variants: RoBell-RVFL targets class-imbalanced datasets via a generalized Bell activation, a dual-model collaborative RVFL addresses conventional RVFL limitations, and an uncertainty-aware ensemble deep RVFL improves on dRVFL/edRVFL by weighting training samples.