Abstract | ||
---|---|---|
The problem of determining an optimal training schedule for a locally recurrent neural network is discussed. Specifically, the proper choice of the most informative measurement data guaranteeing the reliable prediction of the neural network response is considered. Based on a scalar measure of the performance defined on the Fisher information matrix related to the network parameters, the problem was formulated in terms of optimal experimental design. Then, its solution can be readily achieved via the adaptation of effective numerical algorithms based on the convex optimization theory. Finally, some illustrative experiments are provided to verify the presented approach. |
Year | DOI | Venue |
---|---|---|
2009 | 10.1007/978-3-642-04274-4_9 | soft computing |
Keywords | Field | DocType |
neural network,fisher information matrix | Computer science,Scalar (physics),Recurrent neural network,Probabilistic neural network,Fisher information,Artificial intelligence,Artificial neural network,Convex optimization,Machine learning | Conference |
Volume | ISSN | Citations |
5768 | 0302-9743 | 21 |
PageRank | References | Authors |
0.95 | 7 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Krzysztof Patan | 1 | 151 | 18.13 |
Maciej Patan | 2 | 95 | 11.99 |