Nonlinear programming techniques are used to train Elman(1990) type simple recurrent neural networks to reconcile simulated measurements for a simple dynamic system, a draining tank. The networks are trained in a batch mode using the BFGS quasi-Newton nonlinear optimization algorithm. Noisy data are used for both training the networks and testing network performance. Recurrent Elman networks are able to significantly reduce the noise level in the process measurements without explicity knowledge of the nonlinear dynamics of the system.

PDF, PS

In Proceedings of the International Joint Conference on Neural Networks (IJCNN-92), II, 901--906, Baltimore, MD, 1992. Piscataway, NJ: IEEE.

@inproceedings{karjala:ijcnn92, title={Data Rectification Using Recurrent (Elman) Neural Networks}, author={Thomas W. Karjala and David M. Himmelblau and Risto Miikkulainen}, booktitle={Proceedings of the International Joint Conference on Neural Networks (IJCNN-92)}, volume={II}, address={Baltimore, MD}, publisher={Piscataway, NJ: IEEE}, pages={901--906}, url="http://nn.cs.utexas.edu/?karjala:ijcnn92", year={1992} }