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

FindBook

Google Book

Amazon

博客來

Be More With Less : = Scaling Deep-Learning with Minimal Supervision.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Be More With Less :/
其他題名:	Scaling Deep-Learning with Minimal Supervision.
作者:	Wang, Yaqing.
面頁冊數:	1 online resource (159 pages)
附註:	Source: Dissertations Abstracts International, Volume: 85-01, Section: A.
Contained By:	Dissertations Abstracts International85-01A.
標題:	Adaptation. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30505897click for full text (PQDT)
ISBN:	9798379843373

Be More With Less : = Scaling Deep-Learning with Minimal Supervision.
Wang, Yaqing.

Be More With Less :Scaling Deep-Learning with Minimal Supervision. - 1 online resource (159 pages)

Source: Dissertations Abstracts International, Volume: 85-01, Section: A.

Thesis (Ph.D.)--Purdue University, 2022.

Includes bibliographical references

Large-scale deep learning models have reached previously unattainable performance for various tasks. However, the ever-growing resource consumption of neural networks generates large carbon footprint, brings difficulty for academics to engage in research and stops emerging economies from enjoying growing Artificial Intelligence (AI) benefits. To further scale AI to bring more benefits, two major challenges need to be solved. Firstly, even though large-scale deep learning models achieved remarkable success, their performance is still not satisfactory when fine-tuning with only a handful of examples, thereby hindering widespread adoption in real-world applications where a large scale of labeled data is difficult to obtain. Secondly, current machine learning models are still mainly designed for tasks in closed environments where testing datasets are highly similar to training datasets. When the deployed datasets have distribution shift relative to collected training data, we generally observe degraded performance of developed models. How to build adaptable models becomes another critical challenge. To address those challenges, in this dissertation, we focus on two topics: few-shot learning and domain adaptation, where few-shot learning aims to learn tasks with limited labeled data and domain adaption address the discrepancy between training data and testing data. In Part 1, we show our few-shot learning studies. The proposed few-shot solutions are built upon large-scale language models with evolutionary explorations from improving supervision signals, incorporating unlabeled data and improving few-shot learning abilities with lightweight fine-tuning design to reduce deployment costs. In Part 2, domain adaptation studies are introduced. We develop a progressive series of domain adaption approaches to transfer knowledge across domains efficiently to handle distribution shifts, including capturing common patterns across domains, adaptation with weak supervision and adaption to thousands of domains with limited labeled data and unlabeled data.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2023

Mode of access: World Wide Web

ISBN: 9798379843373Subjects--Topical Terms:

3562958
Adaptation.
Index Terms--Genre/Form:

542853
Electronic books.

Be More With Less : = Scaling Deep-Learning with Minimal Supervision.
LDR:03323nmm a2200361K 4500 001 2364835
005 20231212064432.5
006 m o d
007 cr mn ---uuuuu
008 241011s2022 xx obm 000 0 eng d
020 $a 9798379843373
035 $a (MiAaPQ)AAI30505897
035 $a (MiAaPQ)Purdue19678611
035 $a AAI30505897
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Wang, Yaqing. $3 3705656
245 1 0 $a Be More With Less : $b Scaling Deep-Learning with Minimal Supervision.
264 0 $c 2022
300 $a 1 online resource (159 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: 85-01, Section: A.
500 $a Advisor: Gao, Jing.
502 $a Thesis (Ph.D.)--Purdue University, 2022.
504 $a Includes bibliographical references
520 $a Large-scale deep learning models have reached previously unattainable performance for various tasks. However, the ever-growing resource consumption of neural networks generates large carbon footprint, brings difficulty for academics to engage in research and stops emerging economies from enjoying growing Artificial Intelligence (AI) benefits. To further scale AI to bring more benefits, two major challenges need to be solved. Firstly, even though large-scale deep learning models achieved remarkable success, their performance is still not satisfactory when fine-tuning with only a handful of examples, thereby hindering widespread adoption in real-world applications where a large scale of labeled data is difficult to obtain. Secondly, current machine learning models are still mainly designed for tasks in closed environments where testing datasets are highly similar to training datasets. When the deployed datasets have distribution shift relative to collected training data, we generally observe degraded performance of developed models. How to build adaptable models becomes another critical challenge. To address those challenges, in this dissertation, we focus on two topics: few-shot learning and domain adaptation, where few-shot learning aims to learn tasks with limited labeled data and domain adaption address the discrepancy between training data and testing data. In Part 1, we show our few-shot learning studies. The proposed few-shot solutions are built upon large-scale language models with evolutionary explorations from improving supervision signals, incorporating unlabeled data and improving few-shot learning abilities with lightweight fine-tuning design to reduce deployment costs. In Part 2, domain adaptation studies are introduced. We develop a progressive series of domain adaption approaches to transfer knowledge across domains efficiently to handle distribution shifts, including capturing common patterns across domains, adaptation with weak supervision and adaption to thousands of domains with limited labeled data and unlabeled data.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2023
538 $a Mode of access: World Wide Web
650 4 $a Adaptation. $3 3562958
650 4 $a Multilingualism. $3 598147
650 4 $a Sensors. $3 3549539
650 4 $a Bilingual education. $3 2122778
650 4 $a Education. $3 516579
650 4 $a Language. $3 643551
650 4 $a Web studies. $3 2122754
655 7 $a Electronic books. $2 lcsh $3 542853
690 $a 0282
690 $a 0515
690 $a 0679
690 $a 0646
710 2 $a ProQuest Information and Learning Co. $3 783688
710 2 $a Purdue University. $3 1017663
773 0 $t Dissertations Abstracts International $g 85-01A.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30505897 $z click for full text (PQDT)