國立虎尾科技大學 |

Effective and Efficient Continual Learning.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Effective and Efficient Continual Learning./
作者:	Wang, Zifeng.
面頁冊數:	1 online resource (142 pages)
附註:	Source: Dissertations Abstracts International, Volume: 85-03, Section: B.
Contained By:	Dissertations Abstracts International85-03B.
標題:	Computer science. -
電子資源:	click for full text (PQDT)
ISBN:	9798380156295

Effective and Efficient Continual Learning.
Wang, Zifeng.

Effective and Efficient Continual Learning. - 1 online resource (142 pages)

Source: Dissertations Abstracts International, Volume: 85-03, Section: B.

Thesis (Ph.D.)--Northeastern University, 2023.

Includes bibliographical references

Continual Learning (CL) aims to develop models that mimic the human ability to learn continually without forgetting knowledge acquired earlier. While traditional machine learning methods focus on learning with a certain dataset (task), CL methods adapt a single model to learn a sequence of tasks continually.In this thesis, we target developing effective and efficient CL methods under different challenging and resource-limited settings. Specifically, we (1) leverage the idea of sparsity to achieve cost-effective CL, (2) propose a novel prompting-based paradigm for parameter-efficient CL, and (3) utilize task-invariant and task-specific knowledge to enhance existing CL methods in a general way.We first introduce our sparsity-based CL methods. The first method, Learn-Prune-Share (LPS), splits the network into task-specific partitions, leading to no forgetting, while maintaining memory efficiency. Moreover, LPS integrates a novel selective knowledge sharing scheme, enabling adaptive knowledge sharing in an end-to-end fashion. Taking a step further, we present Sparse Continual Learning (SparCL), a novel framework that leverages sparsity to enable cost-effective continual learning on edge devices. SparCL achieves both training acceleration and accuracy preservation through the synergy of three aspects: weight sparsity, data efficiency, and gradient sparsity.Secondly, we present a new paradigm, prompting-based CL, that aims to train a more succinct memory system that is both data and memory efficient. We first propose a method that learns to dynamically prompt (L2P) a pre-trained model to learn tasks sequentially under different task transitions, where prompts are small learnable parameters maintained in a memory space. We then improve L2P by proposing Dual Prompt, which decouples prompts into complementary "General" and "Expert" prompts to learn task-invariant and task-specific instructions, respectively.Finally, we propose DualHSIC, a simple and effective CL method that generalizes the idea of leveraging task-invariant and task-specific knowledge. DualHSIC consists of two complementary components that stem from the so-called Hilbert Schmidt independence criterion (HSIC): HSIC-Bottleneck for Rehearsal (HBR) lessens the inter-task interference and HSIC Alignment (HA) promotes task-invariant knowledge sharing.Comprehensive experimental results demonstrate the effectiveness and efficiency of our methods over the state-of-the-art methods on multiple CL benchmarks.

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

Mode of access: World Wide Web

ISBN: 9798380156295Subjects--Topical Terms:

573171
Computer science.
Subjects--Index Terms:

Continual learningIndex Terms--Genre/Form:

554714
Electronic books.

Effective and Efficient Continual Learning.
LDR:03844ntm a22003977 4500 001 1143861
005 20240517105012.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798380156295
035 $a (MiAaPQ)AAI30638130
035 $a AAI30638130
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Wang, Zifeng. $3 1468665
245 1 0 $a Effective and Efficient Continual Learning.
264 0 $c 2023
300 $a 1 online resource (142 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: 85-03, Section: B.
500 $a Advisor: Dy, Jennifer.
502 $a Thesis (Ph.D.)--Northeastern University, 2023.
504 $a Includes bibliographical references
520 $a Continual Learning (CL) aims to develop models that mimic the human ability to learn continually without forgetting knowledge acquired earlier. While traditional machine learning methods focus on learning with a certain dataset (task), CL methods adapt a single model to learn a sequence of tasks continually.In this thesis, we target developing effective and efficient CL methods under different challenging and resource-limited settings. Specifically, we (1) leverage the idea of sparsity to achieve cost-effective CL, (2) propose a novel prompting-based paradigm for parameter-efficient CL, and (3) utilize task-invariant and task-specific knowledge to enhance existing CL methods in a general way.We first introduce our sparsity-based CL methods. The first method, Learn-Prune-Share (LPS), splits the network into task-specific partitions, leading to no forgetting, while maintaining memory efficiency. Moreover, LPS integrates a novel selective knowledge sharing scheme, enabling adaptive knowledge sharing in an end-to-end fashion. Taking a step further, we present Sparse Continual Learning (SparCL), a novel framework that leverages sparsity to enable cost-effective continual learning on edge devices. SparCL achieves both training acceleration and accuracy preservation through the synergy of three aspects: weight sparsity, data efficiency, and gradient sparsity.Secondly, we present a new paradigm, prompting-based CL, that aims to train a more succinct memory system that is both data and memory efficient. We first propose a method that learns to dynamically prompt (L2P) a pre-trained model to learn tasks sequentially under different task transitions, where prompts are small learnable parameters maintained in a memory space. We then improve L2P by proposing Dual Prompt, which decouples prompts into complementary "General" and "Expert" prompts to learn task-invariant and task-specific instructions, respectively.Finally, we propose DualHSIC, a simple and effective CL method that generalizes the idea of leveraging task-invariant and task-specific knowledge. DualHSIC consists of two complementary components that stem from the so-called Hilbert Schmidt independence criterion (HSIC): HSIC-Bottleneck for Rehearsal (HBR) lessens the inter-task interference and HSIC Alignment (HA) promotes task-invariant knowledge sharing.Comprehensive experimental results demonstrate the effectiveness and efficiency of our methods over the state-of-the-art methods on multiple CL benchmarks.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Computer science. $3 573171
650 4 $a Computer engineering. $3 569006
653 $a Continual learning
653 $a Deep learning
653 $a Machine learning
653 $a Memory efficient
653 $a Inter-task interference
655 7 $a Electronic books. $2 local $3 554714
690 $a 0464
690 $a 0984
690 $a 0800
710 2 $a Northeastern University. $b Electrical and Computer Engineering. $3 1182193
710 2 $a ProQuest Information and Learning Co. $3 1178819
773 0 $t Dissertations Abstracts International $g 85-03B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30638130 $z click for full text (PQDT)