The errors-in-variables cost function for learning neural networks with noisy inputs

摘要

Currently, most learning algorithms for neural network modelling are based on the Output Error approach, using a least squares cost function. This method provides good results when the network is trained with noisy output data, but special care must be taken when training with noisy input data, or when both inputs and outputs contain noise if the deriven NN parameters will ge used with exact inputs, as is the case in simulations or inversing control. This paper introduces a novel cost function for learning NN with noisy inputs, based on the Errors-In-Variables cost function. A learning scheme is presented and examples are given demonstrating the imporved performance in Neural Network curve fitting, at the cost of performing more calculations.