Analysis of time series by re-sampling

• Main Authors: Mohamed, B. I., Elfaki, Faiz Ahmed Mohamed, Daoud, Jamal Ibrahim, Azram, Mohammad
• Format: Conference or Workshop Item
• Language: English
• Published: Kulliyyah of Engineering, IIUM 2012
• Subjects: QA Mathematics
• Online Access: http://irep.iium.edu.my/24644/4/24644_Analysis%20of%20time%20Series%20by%20re-sampling.pdf; http://irep.iium.edu.my/24644/; http://www.iium.edu.my/icmae/12/

The Box-Jenkins methodology is very often used in financier when the time series are analyzed. The estimations of the parameters of the selected models are one of the first tasks of the analysis. The important problem that emerges in connection with the parameters estimation is the problem of their accuracy. This accuracy is often characterized by the bias and standard deviation. When we want to determine these characteristics by the exact methods some problems often emerge. One possibility of the solution of these problems is the bootstrap methods application. Three different approaches of the application of these methods in the autoregressive model are demonstrated in this paper. Simulation studies are conducted to evaluate the methods.