Abstract: In this article, we aim to offer an interpretable learning paradigm for incremental random weight neural networks (IRWNNs). IRWNNs have become a hot research direction of neural network ...