Chaotic time series prediction using phase space reconstruction based conceptor network
- PDF / 1,899,001 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 26 Downloads / 193 Views
(0123456789().,-volV)(0123456789(). ,- volV)
RESEARCH ARTICLE
Chaotic time series prediction using phase space reconstruction based conceptor network Anguo Zhang1,2,3
•
Zheng Xu4
Received: 17 October 2019 / Revised: 30 April 2020 / Accepted: 20 June 2020 Ó Springer Nature B.V. 2020
Abstract The Conceptor network is a new framework of reservoir computing (RC), in addition to the features of easy training, global convergence, it can online learn new classes of input patterns without complete re-learning from all the training data. The conventional connection topology and weights of the hidden layer (reservoir) of RC are initialized randomly, and are fixed to be no longer fine-tuned after initialization. However, it has been demonstrated that the reservoir connection of RC plays an important role in the computational performance of RC. Therefore, in this paper, we optimize the Conceptor’s reservoir connection and propose a phase space reconstruction (PSR) -based reservoir generation method. We tested the generation method on time series prediction task, and the experiment results showed that the proposed PSR-based method can improve the prediction accuracy of Conceptor networks. Further, we compared the PSR-based Conceptor with two Conceptor networks of other typical reservoir topologies (random connected, cortex-like connected), and found that all of their prediction accuracy showed a nonlinear decline trend with increasing storage load, but in comparison, our proposed PSRbased method has the best accuracy under different storage loads. Keywords Conceptor Reservoir computing Phase space reconstruction Time series prediction
Introduction Unlike traditional recurrent neural networks (RNNs), such as Long Short-Term Memory (LSTM), that use error backpropagation through time (BPTT) algorithm to train the network, reservoir computing (RC), another type of RNN, only uses the simple linear regression meethod to train its output weights, while other connection weights (like input weights, hidden weights) are fixed once initialized. Due to the advantages of high accuracy, fast & Anguo Zhang [email protected] 1
College of Physics and Information Engineering, Fuzhou University, Fuzhou 350108, China
2
Key Laboratory of Medical Instrumentation and Pharmaceutical Technology of Fujian Province, Fuzhou 350116, China
3
Research Institute of Ruijie, Ruijie Networks Co., Ltd., Fuzhou 350002, China
4
School of Mechanical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
training, and global convergence (Ding et al. 2005), some typical RC models, such as echo state network (ESN) (Jaeger and Haas 2004) and liquid state machine (LSM) (Maass et al. 2002), have been widely applied to time series prediction (Li et al. 2015; Ma et al. 2009; Jaeger et al. 2007; Li et al. 2016; Wang et al. 2019; Hu et al. 2020), pattern classification (Skowronski and Harris 2007; M E et al. 2009; Hu et al. 2015; Zhang et al. 2019), and anomaly detection (Chen et al. 2018; Mohammadpoory et al. 2019), and so
Data Loading...
