Neural Networks: A Comprehensive Foundation
Neural Networks: A Comprehensive Foundation
Stable Output Feedback in Reservoir Computing Using Ridge Regression
ICANN '08 Proceedings of the 18th international conference on Artificial Neural Networks, Part I
Time-series forecasting using flexible neural tree model
Information Sciences: an International Journal
ICANN'06 Proceedings of the 16th international conference on Artificial Neural Networks - Volume Part I
Noise-Robust Automatic Speech Recognition Using a Predictive Echo State Network
IEEE Transactions on Audio, Speech, and Language Processing
Analog circuit fault diagnosis with echo state networks based on corresponding clusters
ISNN'11 Proceedings of the 8th international conference on Advances in neural networks - Volume Part I
Modular state space of echo state network
Neurocomputing
| Hi-index | 0.01 |
Reservoir computing methods have become popular; however, the nature of the dynamical reservoir (DR) is not thoroughly understood yet. We propose complex echo state network (CESN), the construction process of its DR is determined by five growth factors. The relationships between CESN connectivity structure and its performance are investigated when predicting nonlinear time series. We also introduce a quantifiable characteristic for the connectivity structure-the connectivity index, and a tool to measure the richness of reservoir states-the omega-complexity index. It is demonstrated from the experimental results that connectivity structure of the reservoirs has significant effect on theirs prediction performance, the omega-complexity index can be used as a performance predictor, and particular configuration of the growth factors and corresponding connectivity index can yield optimal performance.