Optimization of Parameters of Echo State Network and Its Application to Underwater Robot

Kazuo Ishii; Tijn Van der Zant; Vlatko Becanovic; Paul Ploger

doi:10.11499/sicep.2004.0_13_5

SICE Annual Conference 2004

DOI https://doi.org/10.11499/sicep.2004.0_13_5

Conference information

Host: The Society of Instrument and Control Engineers

Co-host: The IEEE Industrial Electronics Society, The IEEE Robotics and Automation Society, The IEEE Control Systems Society, The IEEE Systems, Man and Cybernetics Society, The Instrumentation, Systems, and Automation Society

Optimization of Parameters of Echo State Network and Its Application to Underwater Robot

*Kazuo Ishii, Tijn Van der Zant, Vlatko Becanovic, Paul Ploger

Author information

CONFERENCE PROCEEDINGS FREE ACCESS

Pages 13

Details

Abstract

Echo State Networks (ESNs) is a recurrent neural network. Finding a good one depends on choosing the right parameters for generation of the reservoir, intuition and luck. The proposed method eliminates the need for the tuning by hand by replacing it with a double evolutionary computation. First a broad search to find the right parameters. Then a search directly on the connectivity matrices fine-tunes the ESN. Both steps show improvements over other methods for an experimental dataset of underwater robot.

Corresponding author

Register with J-STAGE for free!