A signal processing framework based on dynamic neural networks with application to problems in adaptation, filtering, and classification

By Lee A. Feldkamp, G. Puskorius. 1998

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Deep learning in neural networks: An overview (2014, methodology): “3), the most general variant of this technique uses an RNN instead of an FNN as world model (Schmidhuber, 1990d, 1991c; Feldkamp and Puskorius, 1998).”
Handbook of Neural Network Signal Processing (2000, methodology): “, much higher sparsity than for SVMs can be achieved), yet take less computing time than the RVM [68].”
Adaptive behavior with fixed weights in RNN: an overview (2002, methodology): “The training method used in all the tasks above is based on backpropagation through time (BPTT) and the multistream extended Kalman filter algorithm; see [9] for details.”
Artificial neural network modeling of water table depth fluctuations (2001, methodology): “…technique for practical problems Pearlmutter , 1995] and has been widely used to train RNN to perform various tasks [ Williams and Zipser , 1995; Puskorius et 1996; Feldkamp and Puskorius , 1998; Coulibaly et al ., 2001]. comprehensive description of the truncated BPTT is provided by Haykin [1999].”
Recurrent neural networks for remaining useful life estimation (2008, methodology): “Our RNN, BPTT, and EKF implementation is based on the many papers on this topic by Feldkamp and Puskorious as summarized in [4].”
Simple and conditioned adaptive behavior from Kalman filter trained recurrent networks (2003, methodology): “To facilitate training, we scaled the data; the network input was ypk ¼ 0:01yk and the network target was xpk ¼ 0:1xk: We used 20-stream GEKF training (Feldkamp & Puskorius, 1998).”