Recurrent Spectral Network (RSN): Shaping a discrete map to reach automated classification

Abstract

A novel strategy to automated classification is introduced which exploits a fully trained dynamical system to steer items belonging to different categories towards distinct asymptotic target destinations. These latter are incorporated into the model by taking advantage of the spectral decomposition of the operator that rules the linear evolution across the processing network. Non-linear terms act for a transient and allow to disentangle the data supplied as initial condition to the discrete dynamical system. The system effectively aligns along assigned directions, which reflect the specificity of the provided input and that are encoded in the loss function via suitable spectral projections. The network can be equipped with several memory kernels which can be sequentially activated for serial datasets handling. Our novel approach to classification, that we here term Recurrent Spectral Network (RSN), is successfully challenged against a simple test-bed model, created for illustrative purposes, as well as a standard dataset for image processing training.