Abstract—For the past decades, China’s economy has continued its extraordinary expansion and played increasingly decisive world economic roles. The exchange rate of China’s currency, the renminbi (RMB), has been focused from investors and policymakers all over the world. Consequently, the prediction of the exchange rate of US dollar-CHINA RMB (ER-RMB) has attracted considerable attention in recent years. This study employs certain statistical, soft computing approaches, and their hybrid models to predict the ER-RMB. While the statistical model includes multiple regression (MR), the soft computing models contain Artificial Neural Networks (ANN) and extreme learning machine (ELM). The hybrid modeling schemes include two different combinations of the models. They are the combination of MR and ANN (MR-ANN) and MR and ELM (MR-ELM). The MR component of the hybrid models is established for a selection of fewer explanatory variables, wherein the selected variables are of higher importance. The other components of the hybrid models are then designed to produce forecasts based on those important explanatory variables. In addition, a real dataset of exchange rate of US dollar-CHINA RMB, containing 11 relevant explanatory variables, from July, 2005 to December, 2013 was collected and analyzed. The prediction results reveal that the proposed models, MR-ANN and MR-ELM, are able to accurately predict the ER-RMB. In addition, the proposed MR-ANN model exhibits the best performance among all the forecasting models.

Index Terms—forecast, Renminbi, hybrid, soft computing, statistical model

Cite: Yuehjen E. Shao, Chien-Chin Li, and Po-Yu Chang, "Forecasting the Exchange Rate of US Dollar-China Renminbi Using Hybrid Techniques of Statistical and Soft Computing Approaches," Journal of Industrial and Intelligent Information, Vol. 4, No. 4, pp. 235-240, July 2016. doi: 10.18178/jiii.4.4.235-240