東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Charting Cellular States, One Cell a...

Lopez, Romain.

FindBook

Google Book

Amazon

博客來

Charting Cellular States, One Cell at a Time: Computational, Inferential and Modeling Perspectives.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Charting Cellular States, One Cell at a Time: Computational, Inferential and Modeling Perspectives./
作者:	Lopez, Romain.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2021,
面頁冊數:	198 p.
附註:	Source: Dissertations Abstracts International, Volume: 83-02, Section: B.
Contained By:	Dissertations Abstracts International83-02B.
標題:	Statistics. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28494591
ISBN:	9798535562513

Charting Cellular States, One Cell at a Time: Computational, Inferential and Modeling Perspectives.
Lopez, Romain.

Charting Cellular States, One Cell at a Time: Computational, Inferential and Modeling Perspectives. - Ann Arbor : ProQuest Dissertations & Theses, 2021 - 198 p.

Source: Dissertations Abstracts International, Volume: 83-02, Section: B.

Thesis (Ph.D.)--University of California, Berkeley, 2021.

This item must not be sold to any third party vendors.

In this thesis, we pursue the design of Bayesian models as well as large-scale approximate inference procedures for the analysis of single-cell transcriptomics data. First, we rely on Bayesian statistics to formalize several central algorithmic steps of single-cell transcriptomics data analysis. In particular, we introduce single-cell Variational Inference (scVI), a deep generative model that separates technical noise from biological signal, and is readily applicable de novo cell type identification, batch effect correction, and differential expression. As follow-up work, we introduce single-cell ANnotation using Variational Inference (scANVI), an extension of scVI for sharing annotation of cells across experimental conditions. scANVI is a fully-probabilistic method for harmonizing collections of single-cell RNA sequencing datasets, and therefore provides a principled way of quantifying the uncertainty of cell state annotation. The inference procedure for both methods is based on auto-encoding variational Bayes and scales to millions of cells. In a second line of work, we expose two pathological behaviors of variational inference we encountered while developing scVI. First, the presence of an encoder network leads to entanglement of the latent variables, an undesirable behavior for factoring out unwanted variations from embeddings. To address this problem, we introduce the Hilbert-Schmidt Information Criterion (HSIC) constrained Variational Autoencoder (VAE), a generic variant of the VAE that incorporates a non-parametric measure of dependence as an explicit constraint for the variational distribution. Second, we expose the unsuitability of the variational distribution for Bayesian decision-making and introduce a simple procedure for making more accurate decisions with VAEs. In particular, we explore alternative training procedures, based on using separate variational bounds for learning the generative model and variational posterior.

ISBN: 9798535562513Subjects--Topical Terms:

517247
Statistics.

Charting Cellular States, One Cell at a Time: Computational, Inferential and Modeling Perspectives.
LDR:03017nmm a2200313 4500 001 2282951
005 20211022115807.5
008 220723s2021 ||||||||||||||||| ||eng d
020 $a 9798535562513
035 $a (MiAaPQ)AAI28494591
035 $a AAI28494591
040 $a MiAaPQ $c MiAaPQ
100 1 $a Lopez, Romain. $3 3561837
245 1 0 $a Charting Cellular States, One Cell at a Time: Computational, Inferential and Modeling Perspectives.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2021
300 $a 198 p.
500 $a Source: Dissertations Abstracts International, Volume: 83-02, Section: B.
500 $a Advisor: Jordan, Michael I.;Yosef, Nir.
502 $a Thesis (Ph.D.)--University of California, Berkeley, 2021.
506 $a This item must not be sold to any third party vendors.
520 $a In this thesis, we pursue the design of Bayesian models as well as large-scale approximate inference procedures for the analysis of single-cell transcriptomics data. First, we rely on Bayesian statistics to formalize several central algorithmic steps of single-cell transcriptomics data analysis. In particular, we introduce single-cell Variational Inference (scVI), a deep generative model that separates technical noise from biological signal, and is readily applicable de novo cell type identification, batch effect correction, and differential expression. As follow-up work, we introduce single-cell ANnotation using Variational Inference (scANVI), an extension of scVI for sharing annotation of cells across experimental conditions. scANVI is a fully-probabilistic method for harmonizing collections of single-cell RNA sequencing datasets, and therefore provides a principled way of quantifying the uncertainty of cell state annotation. The inference procedure for both methods is based on auto-encoding variational Bayes and scales to millions of cells. In a second line of work, we expose two pathological behaviors of variational inference we encountered while developing scVI. First, the presence of an encoder network leads to entanglement of the latent variables, an undesirable behavior for factoring out unwanted variations from embeddings. To address this problem, we introduce the Hilbert-Schmidt Information Criterion (HSIC) constrained Variational Autoencoder (VAE), a generic variant of the VAE that incorporates a non-parametric measure of dependence as an explicit constraint for the variational distribution. Second, we expose the unsuitability of the variational distribution for Bayesian decision-making and introduce a simple procedure for making more accurate decisions with VAEs. In particular, we explore alternative training procedures, based on using separate variational bounds for learning the generative model and variational posterior.
590 $a School code: 0028.
650 4 $a Statistics. $3 517247
650 4 $a Cellular biology. $3 3172791
650 4 $a Computer science. $3 523869
650 4 $a Gene expression. $3 643979
650 4 $a Datasets. $3 3541416
650 4 $a Decision making. $3 517204
650 4 $a Algorithms. $3 536374
650 4 $a Annotations. $3 3561780
650 4 $a Visualization. $3 586179
690 $a 0463
690 $a 0379
690 $a 0984
710 2 $a University of California, Berkeley. $b Electrical Engineering & Computer Sciences. $3 1671057
773 0 $t Dissertations Abstracts International $g 83-02B.
790 $a 0028
791 $a Ph.D.
792 $a 2021
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28494591