RMB Exchange Rate Prediction Based on Bayesian
The traditional time series analysis and prediction methods do not take into account the prior information of samples and parameters, resulting in a large deviation between the prediction results and the actual data, and the Bayesian parameter estimation method can make full use of the prior information of the parameters. The variance of the estimated parameters is smaller, the estimated results are more accurate, and the predicted results are more real. In order to correctly analyze and predict the changing trend of RMB exchange rate, this paper selects exchange rate data of 995 working days from August 1, 2015 to August 30, 2019 to model the exchange rate of RMB against US dollar in time series autoregressive model. By using the MCMC method, I carry out the model parameters with Gibbs sampling estimation, which makes the prediction of the model more accurate.
KeywordsAutoregressive model Bayesian parameter estimation MCMC method Gibbs sampling
- 5.Das, M., Ghosh, S.K.: semBnet: a semantic Bayesian network for multivariate prediction of meteorological time series data. Pattern Recogn. Lett. 187, 256–281 (2017)Google Scholar