Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Echo state networks with sparse output connections

: Kobialka, H.-U.; Kayani, U.


Diamantaras, K.:
Artificial neural networks - ICANN 2010. 20th international conference. Pt.1 : Thessaloniki, Greece, September 15-18, 2010; Proceedings
Berlin: Springer, 2010 (Lecture Notes in Computer Science 6352)
ISBN: 3-642-15818-8
ISBN: 978-3-642-15818-6
ISSN: 0302-9743
International Conference on Artificial Neural Networks (ICANN) <20, 2010, Thessaloniki>
Fraunhofer IAIS ()
echo state network; feature selection

An Echo State Network transforms an incoming time series signal into a high-dimensional state space, and, of course, not every dimension may contribute to the solution. We argue that giving low weights via linear regression is not sufficient. Instead irrelevant features should be entirely excluded from directly contributing to the output nodes. We conducted several experiments using two state-of-the-art feature selection algorithms. Results show significant reduction of the generalization error.