SYSTEM AND PROCEDURE OF A MONOTON NEURAL OPERATOR NETWORK TECHNICAL FIELD

摘要

A system for training neural work comprising an input interface for accessing input data for the neural network and a processor in communication with the input interface. The processor is programmed to receive an input to the neural network and to output a trained neural network using forward propagation and backward propagation, the forward propagation including the use of a root finding procedure to identify a fixed point that is one or more parameters of the neural network wherein the back propagation comprises identifying a derivative of loss associated with the parameters of the network.