國立虎尾科技大學 |

Bayesian Deep Learning With Random Feature-Based Gaussian Processes.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	Bayesian Deep Learning With Random Feature-Based Gaussian Processes./
Author:	Liu, Yuhao.
Description:	1 online resource (144 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 84-12, Section: B.
Contained By:	Dissertations Abstracts International84-12B.
Subject:	Statistics. -
Online resource:	click for full text (PQDT)
ISBN:	9798379759261

Bayesian Deep Learning With Random Feature-Based Gaussian Processes.
Liu, Yuhao.

Bayesian Deep Learning With Random Feature-Based Gaussian Processes. - 1 online resource (144 pages)

Source: Dissertations Abstracts International, Volume: 84-12, Section: B.

Thesis (Ph.D.)--State University of New York at Stony Brook, 2023.

Includes bibliographical references

In this thesis, we specifically focus on sparse Gaussian processes (GPs) based on random features (RFs), which offer computational efficiency as they only require matrix multiplication. We investigate the utilization of RF-based deep Gaussian processes (DGPs), in conjunction with analytical filters such as particle filtering (PF), to accommodate sequential learning. Furthermore, we provide a theoretical bound of static regret in the order of O( √ T). In addition, we extend our methods to incorporate ensemble learning for improved stability, as well as dynamic modeling to adapt to time-varying environments. Moreover, we expand our supervised DGPs to unsupervised GP-based deep state-space models (GP-DSSMs) in order to sequentially quantify latent states for real-time tasks.Finally, we apply our proposed models to the problem of graph inference. Specifically, we utilize supervised DGPs to infer edge weights between nodes in signals, while the unsupervised GP-DSSMs are employed to detect possible confounders that may not be directly observed.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798379759261Subjects--Topical Terms:

556824
Statistics.
Subjects--Index Terms:

ConfounderIndex Terms--Genre/Form:

554714
Electronic books.

Bayesian Deep Learning With Random Feature-Based Gaussian Processes.
LDR:02473ntm a22004097 4500 001 1143594
005 20240517104607.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798379759261
035 $a (MiAaPQ)AAI30491280
035 $a AAI30491280
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Liu, Yuhao. $3 1468327
245 1 0 $a Bayesian Deep Learning With Random Feature-Based Gaussian Processes.
264 0 $c 2023
300 $a 1 online resource (144 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 84-12, Section: B.
500 $a Advisor: Zhu, Wei;Djuric, Petar M.;Liu, Ji.
502 $a Thesis (Ph.D.)--State University of New York at Stony Brook, 2023.
504 $a Includes bibliographical references
520 $a In this thesis, we specifically focus on sparse Gaussian processes (GPs) based on random features (RFs), which offer computational efficiency as they only require matrix multiplication. We investigate the utilization of RF-based deep Gaussian processes (DGPs), in conjunction with analytical filters such as particle filtering (PF), to accommodate sequential learning. Furthermore, we provide a theoretical bound of static regret in the order of O( √ T). In addition, we extend our methods to incorporate ensemble learning for improved stability, as well as dynamic modeling to adapt to time-varying environments. Moreover, we expand our supervised DGPs to unsupervised GP-based deep state-space models (GP-DSSMs) in order to sequentially quantify latent states for real-time tasks.Finally, we apply our proposed models to the problem of graph inference. Specifically, we utilize supervised DGPs to infer edge weights between nodes in signals, while the unsupervised GP-DSSMs are employed to detect possible confounders that may not be directly observed.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Statistics. $3 556824
650 4 $a Electrical engineering. $3 596380
650 4 $a Computer engineering. $3 569006
650 4 $a Applied mathematics. $3 1069907
653 $a Confounder
653 $a Dynamic graph
653 $a Gaussian process
653 $a Random feature
653 $a State-space model
655 7 $a Electronic books. $2 local $3 554714
690 $a 0463
690 $a 0544
690 $a 0464
690 $a 0364
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a State University of New York at Stony Brook. $b Applied Mathematics and Statistics. $3 1179106
773 0 $t Dissertations Abstracts International $g 84-12B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30491280 $z click for full text (PQDT)