The Statistical Property of the Wavelet-Based Estimator for Long-Memory Processes with Missing Data

2016 Fiscal Year Final Research Report

• Project/Area Number: 15H06525
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Single-year Grants
• Research Field: Economic statistics
• Research Institution: Aomori Public College
• Principal Investigator: NANAMIYA Kei 青森公立大学, 経営経済学部, 講師 (20755714)
• Project Period (FY): 2015-08-28 – 2017-03-31
• Keywords: 時系列解析 / 長期記憶過程 / 欠測値 / ウェーブレット解析

Outline of Final Research Achievements

We introduce the wavelet-based approximated maximum likelihood estimators for the parameters of the long-memory processes with missing data, and explore the statistical property of these estimators. In our study, by approximating the wavelet coefficients of the processes with missing data as white noise processes in each level and by removing the coefficients including missing data, we propose the wavelet-based approximated maximum likelihood estimators. We show that these estimators are not affected by missing data and have the same statistical property as the wavelet-based approximated maximum likelihood estimators for long-memory processes which are complete data.

Free Research Field

経済時系列解析