Extraction and Forecasting of Financial Returns Based on Hybrid Method of EMD and STSA
LI Xiang-fe · Systems Engineering · 2014
This paper puts forward an improved method by combined EMD and STSA together direct at the defect of STSA method in financial time series analysis.Taking the return data of 6different stock indexes as the research sample and using EMD decomposition method extracted a series of signal components which reflect different time scale information of the original sequence.By using STSA to each component,we have found the reason leading to the diverse changing pattern of the original sequence.Based on this,we propose a method that using the single changing pattern component to estimate the original sequence,the condition and limited range are given as well.The experimental results show that the proposed approach has uniquely superiority in extract the information of time series,and is of high precision and practicability as well.