Font Size: a A A

The Short-term Prediction Research On Exchange Rate Of USD/CNY

Posted on:2017-05-23Degree:MasterType:Thesis
Country:ChinaCandidate:Z L XuFull Text:PDF
GTID:2279330503966663Subject:statistics
Abstract/Summary:PDF Full Text Request
The exchange rate of USD/CNY is an important index of economic relations between China and America, the Central parity rate has been changing since the exchange rate reform,and the reasons is various and complex. To grasp the change of exchange rate information in time, we need to do short-term prediction of exchange rate time series providing decision-making basisfor the country and the enterprise or individual investment and trade activities in the international market. Information granulation has certain advantage in accessing to diverse information, support vector machine(SVM) has also been quite mature in the treatment of nonlinear regression prediction problems, and based on information granulation of support vector machine(SVM) has been widely used in financial time series prediction field.This paper adopts the fuzzy information granulation of support vector machine(SVM) model for short-term prediction research on exchange rate of USD/CNY, it selects the data of USD/CNY on January 4, 2011 to January 25, 2016 as sample.First,We deal the sample data with fuzzy information granulation and getthree fuzzy particles Low, R,Up, the three fuzzy particlesrepresentthe original data changes of minimum, average and maximum;Second, the fuzzy particle data are normalized processing; Third, we train the fuzzy particle of normalizationusing cross validation method, and obtain the optimal parameters of support vector machine(SVM) model;Finally, we use support vector machine(SVM) to forecast the three fuzzy particle data respectively, get threemodelsand the goodness-of-fit are 98.17%, 99.03%, 98.71%.We get the prediction result is that the central parity rate range in the next five days of is(6.5507, 6.5507),and the change trend of exchange rate in the next five days is falling.Upon examination, these are consistent with the actual situation.The results show that the fuzzy information granulation of support vector machine(SVM) modelapplied to the short-term prediction research on exchange rate of USD/CNY is good and robust, and it will have a certain application prospect.
Keywords/Search Tags:Fuzzy information granulation, Support vector machine(SVM), Short-term exchange rate prediction, The change scope and trend
PDF Full Text Request
Related items