On fast supervised learning for normal mixture models with missing information
碩士 === 東海大學 === 統計學系 === 93 === It is an important research issue to deal with mixture models when missing values occur in the data. In this paper, computational strategies using auxiliary indicator matrices are introduced for handling mixtures of multivariate normal distributions in a more efficien...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2005
|
Online Access: | http://ndltd.ncl.edu.tw/handle/14446657169844690511 |
id |
ndltd-TW-093THU00337006 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-093THU003370062016-06-10T04:16:00Z http://ndltd.ncl.edu.tw/handle/14446657169844690511 On fast supervised learning for normal mixture models with missing information 具遺失訊息下多變量混合常態模型之快速監督學習 Hsiu J. Ho 何秀榮 碩士 東海大學 統計學系 93 It is an important research issue to deal with mixture models when missing values occur in the data. In this paper, computational strategies using auxiliary indicator matrices are introduced for handling mixtures of multivariate normal distributions in a more efficient manner, assuming that patterns of missingness are arbitrary and missing at random. We develop a novelly structured EM algorithm which can dramatically save computation time and be exploited in many applications, such as density estimation, supervised clustering and prediction of missing values. In the aspect of multiple imputations for missing data, we also offer a data augmentation scheme using the Gibbs sampler. Our proposed methodologies are illustrated through some real data sets with varying proportions of missing values. Tsung I. Lin 林宗儀 2005 學位論文 ; thesis 27 en_US |
collection |
NDLTD |
language |
en_US |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 東海大學 === 統計學系 === 93 === It is an important research issue to deal with mixture models when missing values occur in the data. In this paper, computational strategies using auxiliary indicator matrices are introduced for handling mixtures of multivariate normal distributions in a more efficient manner, assuming that patterns of missingness are arbitrary and missing at random. We develop a novelly structured EM algorithm which can dramatically save computation time and be exploited in many applications, such as density estimation, supervised clustering and prediction of missing values. In the aspect of multiple imputations for missing data, we also offer a data augmentation scheme using the Gibbs sampler. Our proposed methodologies are illustrated through some real data sets with varying proportions of missing values.
|
author2 |
Tsung I. Lin |
author_facet |
Tsung I. Lin Hsiu J. Ho 何秀榮 |
author |
Hsiu J. Ho 何秀榮 |
spellingShingle |
Hsiu J. Ho 何秀榮 On fast supervised learning for normal mixture models with missing information |
author_sort |
Hsiu J. Ho |
title |
On fast supervised learning for normal mixture models with missing information |
title_short |
On fast supervised learning for normal mixture models with missing information |
title_full |
On fast supervised learning for normal mixture models with missing information |
title_fullStr |
On fast supervised learning for normal mixture models with missing information |
title_full_unstemmed |
On fast supervised learning for normal mixture models with missing information |
title_sort |
on fast supervised learning for normal mixture models with missing information |
publishDate |
2005 |
url |
http://ndltd.ncl.edu.tw/handle/14446657169844690511 |
work_keys_str_mv |
AT hsiujho onfastsupervisedlearningfornormalmixturemodelswithmissinginformation AT héxiùróng onfastsupervisedlearningfornormalmixturemodelswithmissinginformation AT hsiujho jùyíshīxùnxīxiàduōbiànliànghùnhéchángtàimóxíngzhīkuàisùjiāndūxuéxí AT héxiùróng jùyíshīxùnxīxiàduōbiànliànghùnhéchángtàimóxíngzhīkuàisùjiāndūxuéxí |
_version_ |
1718301070035255296 |