東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Statistical and Machine Learning Met...

Zeng, Li.

Linked to FindBook

Google Book

Amazon

博客來

Statistical and Machine Learning Methods in Cancer Genomic Data Analysis.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Statistical and Machine Learning Methods in Cancer Genomic Data Analysis./
Author:	Zeng, Li.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2018,
Description:	130 p.
Notes:	Source: Dissertations Abstracts International, Volume: 80-09, Section: B.
Contained By:	Dissertations Abstracts International80-09B.
Subject:	Biostatistics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13851107
ISBN:	9780438981485

Statistical and Machine Learning Methods in Cancer Genomic Data Analysis.
Zeng, Li.

Statistical and Machine Learning Methods in Cancer Genomic Data Analysis. - Ann Arbor : ProQuest Dissertations & Theses, 2018 - 130 p.

Source: Dissertations Abstracts International, Volume: 80-09, Section: B.

Thesis (Ph.D.)--Yale University, 2018.

This item must not be added to any third party search indexes.

Over the years, statistical and machine learning methods have greatly benefited scientific discovery and understanding through analysis of complex and mounting biological data. Despite the success achieved, many of these methods are not tailored to specific datasets. It would be more effective to develop and apply methods that can incorporate knowledge of the data generation process and the underlying data structure. In this thesis, we present three projects, where we propose novel methods motivated by prior biological knowledge. In Chapter 1, we propose a Bayesian hierarchical model to study the evolution of tumor inside an individual patient, using Whole Genome Sequencing data. Our method is able to identify subgroups of tumor cells with distinct genetic alteration patterns, and infer the phylogenetic structure among them. Results from such analysis lead to insights on tumor heterogeneity, which is important for drug resistance and metastasis. In Chapters 2 and 3, we focus on the analysis of cancer genomic datasets, where both clinical features and gene expression profiles are collected from cancer patients. In Chapter 2, we propose a novel boosting model for sample classification, which is capable of utilizing pathway information to improve prediction performance. The pathway information is incorporated by constructing pathway-based base learners. Improved prediction accuracy is demonstrated through its applications to predict several clinical traits, such as tumor grade, stage, and metastasis status. In Chapter 3, we further generalize the boosting model and unify classification, regression, and survival analysis under the same framework. The new boosting model is thus able to handle categorical, continuous, and survival outcome variables, with different specifications in loss functions. We demonstrate the usefulness of the model in applications to predict drug responses and patients' survival.

ISBN: 9780438981485Subjects--Topical Terms:

1002712
Biostatistics.
Subjects--Index Terms:

Gradient Boosting

Statistical and Machine Learning Methods in Cancer Genomic Data Analysis.
LDR:03191nmm a2200385 4500 001 2272448
005 20201105110057.5
008 220629s2018 ||||||||||||||||| ||eng d
020 $a 9780438981485
035 $a (MiAaPQ)AAI13851107
035 $a AAI13851107
040 $a MiAaPQ $c MiAaPQ
100 1 $a Zeng, Li. $3 1043794
245 1 0 $a Statistical and Machine Learning Methods in Cancer Genomic Data Analysis.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2018
300 $a 130 p.
500 $a Source: Dissertations Abstracts International, Volume: 80-09, Section: B.
500 $a Publisher info.: Dissertation/Thesis.
500 $a Advisor: Zhao, Hongyu.
502 $a Thesis (Ph.D.)--Yale University, 2018.
506 $a This item must not be added to any third party search indexes.
506 $a This item must not be sold to any third party vendors.
520 $a Over the years, statistical and machine learning methods have greatly benefited scientific discovery and understanding through analysis of complex and mounting biological data. Despite the success achieved, many of these methods are not tailored to specific datasets. It would be more effective to develop and apply methods that can incorporate knowledge of the data generation process and the underlying data structure. In this thesis, we present three projects, where we propose novel methods motivated by prior biological knowledge. In Chapter 1, we propose a Bayesian hierarchical model to study the evolution of tumor inside an individual patient, using Whole Genome Sequencing data. Our method is able to identify subgroups of tumor cells with distinct genetic alteration patterns, and infer the phylogenetic structure among them. Results from such analysis lead to insights on tumor heterogeneity, which is important for drug resistance and metastasis. In Chapters 2 and 3, we focus on the analysis of cancer genomic datasets, where both clinical features and gene expression profiles are collected from cancer patients. In Chapter 2, we propose a novel boosting model for sample classification, which is capable of utilizing pathway information to improve prediction performance. The pathway information is incorporated by constructing pathway-based base learners. Improved prediction accuracy is demonstrated through its applications to predict several clinical traits, such as tumor grade, stage, and metastasis status. In Chapter 3, we further generalize the boosting model and unify classification, regression, and survival analysis under the same framework. The new boosting model is thus able to handle categorical, continuous, and survival outcome variables, with different specifications in loss functions. We demonstrate the usefulness of the model in applications to predict drug responses and patients' survival.
590 $a School code: 0265.
650 4 $a Biostatistics. $3 1002712
650 4 $a Genetics. $3 530508
650 4 $a Statistics. $3 517247
653 $a Gradient Boosting
653 $a Pathway-Based Model
653 $a Tumor Heterogeneity
653 $a Tumor Subclone
690 $a 0308
690 $a 0369
690 $a 0463
710 2 $a Yale University. $3 515640
773 0 $t Dissertations Abstracts International $g 80-09B.
790 $a 0265
791 $a Ph.D.
792 $a 2018
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13851107