Abstract
The time series of gold price has the characteristics of nonlinearity, non-stationary and high noise, which cannot be fully described by using traditional single model. Based on the decomposition reconstruction theory, a combined prediction model based on the pole symmetry mode decomposition (ESMD) is proposed. Firstly, the gold price time series is decomposed into multiple eigenmode components by using the ESMD method, and these components are recombined into high, medium and low frequency parts. Secondly, the appropriate prediction methods are selected for the three parts of the different frequencies of recombination: The nonlinear autoregressive neural network predicts the high frequency part, and the IF and low frequency parts are predicted by the multi-task, and least squares support vector machine. Finally, the prediction results of the three parts are integrated to obtain the prediction result of the gold price. It shows that the model proposed in this paper has higher accuracy and is a more effective method for predicting gold price.
Export citation and abstract BibTeX RIS
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.