Echo State Network Optimization using Hybrid- Structure Based Gravitational Search Algorithm

[摘要] The Echo-State Network (ESN) is a robust recurrent neural network and a generalized form of classical neuralnetworks in time-series model designs. ESN inherits a simple approach for training and demonstrates the high computationalcapability to solve non-linear problems. However, input weights and the reservoir's internal weights are pre-defined whenoptimizing with only the output weight matrix. This paper proposes a Hybrid Gravitational Search Algorithm (HGSA) to computeESN output weights. In Gravitational Search Algorithm (GSA), Square Quadratic Programming (SQP) is united as a local searchstrategy to raise the standard GSA algorithm's efficiency. Later, an HGSA-SQP and the validation data set to establish therelation configuration of the ESN output weights. Experimental results indicate that the proposed configuration of HGSA-SQP- ESN is more efficient than the other conventional models of ESN with the minimum generalization error.

[学科分类] 计算机科学（综合）

[关键词] Echo state network;hybrid gravitational search algorithm;network configuration optimization;time seriesprediction