x

## 留言板

Multiple clusters echo state network for chaotic time series prediction

## Multiple clusters echo state network for chaotic time series prediction

Song Qing-Song, Feng Zu-Ren, Li Ren-Hou
• #### Abstract

The chaotic time series prediction problem is considered. A novel type of cortex-like neural network model, i.e. multi-clusters echo state network model （MCESN）, regulated by a group of five growth-factors, is proposed. It is shown that characters of MCESN’ topology can be effectively determined by the growth-factors group; and that it is the MCESN possessing both small-world and scale-free properties of complex network that corresponds to the better prediction performance. In addition, Monte Carlo simulation experiments show that MCESN not only can be trained by easy algorithm, but also can achieve higher accuracy and less standard deviation prediction results than classical echo state networks.

• Funds:

#### Cited By

•  [1] . Diagnosis of capacitively coupled plasma driven by pulse-modulated 27.12 MHz by using an emissive probe. Acta Physica Sinica, 2020, (): . doi: 10.7498/aps.69.20191864 [2] Zhuang Zhi-Ben, Li Jun, Liu Jing-Yi, Chen Shi-Qiang. Image encryption algorithm based on new five-dimensional multi-ring multi-wing hyperchaotic system. Acta Physica Sinica, 2020, 69(4): 040502. doi: 10.7498/aps.69.20191342 [3] Huang Yong-Feng, Cao Huai-Xin, Wang Wen-Hua. Conjugate linear symmetry and its application to \begin{document}${\mathcal{P}}{\mathcal{T}}$\end{document}-symmetry quantum theory. Acta Physica Sinica, 2020, 69(3): 030301. doi: 10.7498/aps.69.20191173 [4] Ren Xian-Li, Zhang Wei-Wei, Wu Xiao-Yong, Wu Lu, Wang Yue-Xia. Prediction of short range order in high-entropy alloys and its effect on the electronic, magnetic and mechanical properties. Acta Physica Sinica, 2020, 69(4): 046102. doi: 10.7498/aps.69.20191671 [5] Liao Tian-Jun, Lü Yi-Xiang. Thermodynamic limit and optimal performance prediction of thermophotovoltaic energy conversion devices. Acta Physica Sinica, 2020, 69(5): 057202. doi: 10.7498/aps.69.20191835
•  Citation:
##### Metrics
• Abstract views:  3591
• Cited By: 0
##### Publishing process
• Received Date:  07 January 2008
• Accepted Date:  09 December 2008
• Published Online:  20 July 2009

## Multiple clusters echo state network for chaotic time series prediction

• 1. 西安交通大学系统工程研究所，西安 710049

Abstract: The chaotic time series prediction problem is considered. A novel type of cortex-like neural network model, i.e. multi-clusters echo state network model （MCESN）, regulated by a group of five growth-factors, is proposed. It is shown that characters of MCESN’ topology can be effectively determined by the growth-factors group; and that it is the MCESN possessing both small-world and scale-free properties of complex network that corresponds to the better prediction performance. In addition, Monte Carlo simulation experiments show that MCESN not only can be trained by easy algorithm, but also can achieve higher accuracy and less standard deviation prediction results than classical echo state networks.

/