Font Size: a A A

The Short-term Prediction Of RMB Exchange Rate Based On Radial Basis Function Network Hybrid Model

Posted on:2021-05-21Degree:MasterType:Thesis
Country:ChinaCandidate:Y LiFull Text:PDF
GTID:2480306311984579Subject:Applied Statistics
Abstract/Summary:PDF Full Text Request
In August 2015,the central bank issued a new exchange rate reform policy that no longer interferes with the daily price of the central parity of the RMB against the US dollar.Subsequently,in October 2016,the renminbi was included in the SDR,which became more and more important in the global economy.The rise of the international status has made the renminbi exchange rate attract worldwide attention.At the same time,since the exchange rate has an impact on the country's formulation of different import and export strategies and corporate hedging for corporate financing,accurate exchange rate prediction is particularly important,and the exchange rate prediction has also attracted the attention of scholars from all walks of life.In this paper,the central parity rate of RMB/US is used as empirical dat a.In order to improve the accuracy of the exchange rate forecast,the CEEMD-DE-GWO-RBF combination model is used.First,the optimized complementar y set empirical mode decomposition method is used to denoise the original ex change rate data to obtain a set of time series data that can be used for pred iction.Then the DE optimization algorithm and the GWO optimization algorit hm are used to optimize the coefficients of the RBF neural network for time series prediction.In addition,in order to measure the prediction accuracy of t he model,three error indicators commonly used to evaluate the prediction acc uracy of the time series model are used.The prediction accuracy is compared with the prediction results of other models through the evaluation indicators.In other models,the CEEMD denoising method is superior to the commonly used EMD and EEMD denoising methods,and the DE-GWO new combination optimization algorithm used in this paper is more effective in global optimiza tion than other non-combination optimization algorithms.Finally,the proposed model and the comparison model are tested by DM to ensure the validity of the model.On the basis of verifying that the RMB/USD exchange rate data can achieve better prediction results,the RMB/EUR exchange rate data is used for further verification.The structure of the article is divided into five p arts.Chapter 1 is the introduction that introducing the background and signific ance of the exchange rate forecast;Chapter 2 is the literature review,which s orts out the domestic and foreign literature;Chapter 3 is the theoretical basis,which describes the denoising model and optimization principles of algorithm s and prediction models;Chapter 4 builds exchange rate prediction models an d conducts empirical analysis of the exchange rate data of RMB/US and RMB/EURO;Chapter 5 is summarizes and prospects.The model proposed in this paper has good prediction performance and is more perfect in the theoretical basis of the model.The prediction results have very important practical significance for the decision-making of the regulatory layer and the investment choices for investors.
Keywords/Search Tags:Exchange rate prediction, combined prediction model, differential gray wolf optimization algorithm, neural network
PDF Full Text Request
Related items