L1 Linear Interpolator of Missing Values in Time Series

We propose a minimum mean absolute error linear interpolator (MMAELI), based on the L1 approach. A linear functional of the observed time series due to non-normal innovations is derived. The solution equation for the coefficients of this linear functional is established in terms of the innovation se...

Full description

Bibliographic Details
Main Authors: Lu, Zudi, Hui, Y.
Format: Journal Article
Published: Kluwer Academic Publishers 2003
Subjects:
Online Access:http://hdl.handle.net/20.500.11937/44023
_version_ 1848756879033892864
author Lu, Zudi
Hui, Y.
author_facet Lu, Zudi
Hui, Y.
author_sort Lu, Zudi
building Curtin Institutional Repository
collection Online Access
description We propose a minimum mean absolute error linear interpolator (MMAELI), based on the L1 approach. A linear functional of the observed time series due to non-normal innovations is derived. The solution equation for the coefficients of this linear functional is established in terms of the innovation series. It is found that information implied in the innovation series is useful for the interpolation of missing values. The MMAELIs of the AR(1) model with innovations following mixed normal and t distributions are studied in detail. The MMAELI also approximates the minimum mean squared error linear interpolator (MMSELI) well in mean squared error but outperforms the MMSELI in mean absolute error. An applicationto a real series is presented. Extensions to the general ARMA model and other time series models are discussed.
first_indexed 2025-11-14T09:19:12Z
format Journal Article
id curtin-20.500.11937-44023
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T09:19:12Z
publishDate 2003
publisher Kluwer Academic Publishers
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-440232017-09-13T16:05:30Z L1 Linear Interpolator of Missing Values in Time Series Lu, Zudi Hui, Y. Autoregressive process missing values minimum mean absolute error linear interpolation innovation departure We propose a minimum mean absolute error linear interpolator (MMAELI), based on the L1 approach. A linear functional of the observed time series due to non-normal innovations is derived. The solution equation for the coefficients of this linear functional is established in terms of the innovation series. It is found that information implied in the innovation series is useful for the interpolation of missing values. The MMAELIs of the AR(1) model with innovations following mixed normal and t distributions are studied in detail. The MMAELI also approximates the minimum mean squared error linear interpolator (MMSELI) well in mean squared error but outperforms the MMSELI in mean absolute error. An applicationto a real series is presented. Extensions to the general ARMA model and other time series models are discussed. 2003 Journal Article http://hdl.handle.net/20.500.11937/44023 10.1007/BF02530494 Kluwer Academic Publishers restricted
spellingShingle Autoregressive process
missing values
minimum mean absolute error
linear interpolation
innovation departure
Lu, Zudi
Hui, Y.
L1 Linear Interpolator of Missing Values in Time Series
title L1 Linear Interpolator of Missing Values in Time Series
title_full L1 Linear Interpolator of Missing Values in Time Series
title_fullStr L1 Linear Interpolator of Missing Values in Time Series
title_full_unstemmed L1 Linear Interpolator of Missing Values in Time Series
title_short L1 Linear Interpolator of Missing Values in Time Series
title_sort l1 linear interpolator of missing values in time series
topic Autoregressive process
missing values
minimum mean absolute error
linear interpolation
innovation departure
url http://hdl.handle.net/20.500.11937/44023