The artificial intelligence method represented by neural network is widely used in the field of nonlinear system estimation. In order to solve the problems that the hidden nodes of neural network are difficult to determine and the calculation is complex, Jaeger proposed an echo state network model for nonlinear system prediction. Aiming at the difficulty of multi-parameter selection of echo state network for nonlinear system prediction, the genetic algorithm and cross-validation method are used to optimize the network parameters. The simulation show that the genetic algorithm optimizes parameters of the echo state network to prediction accuracy is better than the cross-validation. The cross-validation is superior to the genetic algorithm in optimizing efficiency.
Paper
Full text
Research for Parameters Optimization of Echo State Network
Semantic Scholar · Computer Science · 2018
Abstract
The artificial intelligence method represented by neural network is widely used in the field of nonlinear system estimation. In order to solve the problems that the hidden nodes of neural network are difficult to determine and the calculation is complex, Jaeger proposed an echo state network model for nonlinear system prediction. Aiming at the difficulty of multi-parameter selection of echo state network for nonlinear system prediction, the genetic algorithm and cross-validation method are used to optimize the network parameters. The simulation show that the genetic algorithm optimizes parameters of the echo state network to prediction accuracy is better than the cross-validation. The cross-validation is superior to the genetic algorithm in optimizing efficiency.