Effect of Regular and Irregular Missing Data on the Correlation Integral Analysis of Real-Time Series

Md. Mostafizur Rahman KHAN  Noboru TANIZUKA  

Publication
IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E88-A   No.11   pp.3161-3168
Publication Date: 2005/11/01
Online ISSN: 
DOI: 10.1093/ietfec/e88-a.11.3161
Print ISSN: 0916-8508
Type of Manuscript: PAPER
Category: Nonlinear Problems
Keyword: 
time series analysis,  missing data,  correlation integral,  exchange rate,  quasar radio wave intensity,  

Full Text: PDF(413.6KB)>>
Buy this Article




Summary: 
Missing data which inevitably occurs in observed time series may lead to an erroneous result based on the correlation integral analysis. Effects of data, missing at regular and irregular times, on the analyzed result are estimated. A model estimation is obtained for the Lorenz time series. The effects of the missing data in economic and astronomical time series are estimated using the correlation integral analysis. A convenient method of choosing a time lag is proposed to minimize the effect of regularly missing data.