Task and instance quadratic ordering for active online multitask learning

Jing Zhao; Shaoning Pang; Iman Tabatabaei Ardekani; Yuji Sekiya; Daisuke Miyamoto

doi:10.1007/978-3-030-04182-3_38

Task and instance quadratic ordering for active online multitask learning

Jing Zhao^*, Shaoning Pang, Iman Tabatabaei Ardekani, Yuji Sekiya, Daisuke Miyamoto

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding contribution › peer-review

Abstract

For online multitask learning (oMTL), when a chunk of tasks consisting of multiple related instances is received in one batch, the learner normally has the chance to actively order these tasks to improve the learning efficiency. This paper proposes a quadratic ordering method for active oMTL, where instance ordering is integrated into task ordering by taking each instance in one task. The proposed task and instance quadratic ordering is able to facilitate oMTL better than single task ordering. The orderings derived in this paper can be incorporated into any individual oMTL algorithms for active oMTL. The performance evaluations on four real-word datasets demonstrate the benefits of the proposed algorithms.

Original language	English
Title of host publication	Neural Information Processing
Subtitle of host publication	25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13–16, 2018, proceedings, part III
Editors	Long Cheng, Andrew Chi Sing Leung, Seiichi Ozawa
Place of Publication	Cham
Publisher	Springer, Springer Nature
Pages	435-447
Number of pages	13
ISBN (Electronic)	9783030041823
ISBN (Print)	9783030041816
DOIs	https://doi.org/10.1007/978-3-030-04182-3_38
Publication status	Published - 2018
Externally published	Yes
Event	25th International Conference on Neural Information Processing, ICONIP 2018 - Siem Reap, Cambodia Duration: 13 Dec 2018 → 16 Dec 2018

Conference

Conference	25th International Conference on Neural Information Processing, ICONIP 2018
Country/Territory	Cambodia
City	Siem Reap
Period	13/12/18 → 16/12/18

Keywords

Online Multitask Learning (oMTL)
Active oMTL Quadratic ordering
Task ordering
Instance ordering

Access to Document

10.1007/978-3-030-04182-3_38

Cite this

Zhao, J., Pang, S., Ardekani, I. T., Sekiya, Y., & Miyamoto, D. (2018). Task and instance quadratic ordering for active online multitask learning. In L. Cheng, A. C. S. Leung, & S. Ozawa (Eds.), Neural Information Processing: 25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13–16, 2018, proceedings, part III (pp. 435-447). Springer, Springer Nature. https://doi.org/10.1007/978-3-030-04182-3_38

Zhao, Jing ; Pang, Shaoning ; Ardekani, Iman Tabatabaei et al. / Task and instance quadratic ordering for active online multitask learning. Neural Information Processing: 25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13–16, 2018, proceedings, part III. editor / Long Cheng ; Andrew Chi Sing Leung ; Seiichi Ozawa. Cham : Springer, Springer Nature, 2018. pp. 435-447

@inproceedings{debaf0f139844f25924a95c3e1fdd584,

title = "Task and instance quadratic ordering for active online multitask learning",

abstract = "For online multitask learning (oMTL), when a chunk of tasks consisting of multiple related instances is received in one batch, the learner normally has the chance to actively order these tasks to improve the learning efficiency. This paper proposes a quadratic ordering method for active oMTL, where instance ordering is integrated into task ordering by taking each instance in one task. The proposed task and instance quadratic ordering is able to facilitate oMTL better than single task ordering. The orderings derived in this paper can be incorporated into any individual oMTL algorithms for active oMTL. The performance evaluations on four real-word datasets demonstrate the benefits of the proposed algorithms.",

keywords = "Online Multitask Learning (oMTL), Active oMTL Quadratic ordering, Task ordering, Instance ordering",

author = "Jing Zhao and Shaoning Pang and Ardekani, {Iman Tabatabaei} and Yuji Sekiya and Daisuke Miyamoto",

year = "2018",

doi = "10.1007/978-3-030-04182-3_38",

language = "English",

isbn = "9783030041816",

pages = "435--447",

editor = "Long Cheng and Leung, {Andrew Chi Sing} and Seiichi Ozawa",

booktitle = "Neural Information Processing",

publisher = "Springer, Springer Nature",

address = "United States",

note = "25th International Conference on Neural Information Processing, ICONIP 2018 ; Conference date: 13-12-2018 Through 16-12-2018",

}

Zhao, J, Pang, S, Ardekani, IT, Sekiya, Y & Miyamoto, D 2018, Task and instance quadratic ordering for active online multitask learning. in L Cheng, ACS Leung & S Ozawa (eds), Neural Information Processing: 25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13–16, 2018, proceedings, part III. Springer, Springer Nature, Cham, pp. 435-447, 25th International Conference on Neural Information Processing, ICONIP 2018, Siem Reap, Cambodia, 13/12/18. https://doi.org/10.1007/978-3-030-04182-3_38

Task and instance quadratic ordering for active online multitask learning. / Zhao, Jing; Pang, Shaoning; Ardekani, Iman Tabatabaei et al.
Neural Information Processing: 25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13–16, 2018, proceedings, part III. ed. / Long Cheng; Andrew Chi Sing Leung; Seiichi Ozawa. Cham: Springer, Springer Nature, 2018. p. 435-447.

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding contribution › peer-review

TY - GEN

T1 - Task and instance quadratic ordering for active online multitask learning

AU - Zhao, Jing

AU - Pang, Shaoning

AU - Ardekani, Iman Tabatabaei

AU - Sekiya, Yuji

AU - Miyamoto, Daisuke

PY - 2018

Y1 - 2018

N2 - For online multitask learning (oMTL), when a chunk of tasks consisting of multiple related instances is received in one batch, the learner normally has the chance to actively order these tasks to improve the learning efficiency. This paper proposes a quadratic ordering method for active oMTL, where instance ordering is integrated into task ordering by taking each instance in one task. The proposed task and instance quadratic ordering is able to facilitate oMTL better than single task ordering. The orderings derived in this paper can be incorporated into any individual oMTL algorithms for active oMTL. The performance evaluations on four real-word datasets demonstrate the benefits of the proposed algorithms.

AB - For online multitask learning (oMTL), when a chunk of tasks consisting of multiple related instances is received in one batch, the learner normally has the chance to actively order these tasks to improve the learning efficiency. This paper proposes a quadratic ordering method for active oMTL, where instance ordering is integrated into task ordering by taking each instance in one task. The proposed task and instance quadratic ordering is able to facilitate oMTL better than single task ordering. The orderings derived in this paper can be incorporated into any individual oMTL algorithms for active oMTL. The performance evaluations on four real-word datasets demonstrate the benefits of the proposed algorithms.

KW - Online Multitask Learning (oMTL)

KW - Active oMTL Quadratic ordering

KW - Task ordering

KW - Instance ordering

UR - http://www.scopus.com/inward/record.url?scp=85059001505&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-04182-3_38

DO - 10.1007/978-3-030-04182-3_38

M3 - Conference proceeding contribution

SN - 9783030041816

SP - 435

EP - 447

BT - Neural Information Processing

A2 - Cheng, Long

A2 - Leung, Andrew Chi Sing

A2 - Ozawa, Seiichi

PB - Springer, Springer Nature

CY - Cham

T2 - 25th International Conference on Neural Information Processing, ICONIP 2018

Y2 - 13 December 2018 through 16 December 2018

ER -

Zhao J, Pang S, Ardekani IT, Sekiya Y, Miyamoto D. Task and instance quadratic ordering for active online multitask learning. In Cheng L, Leung ACS, Ozawa S, editors, Neural Information Processing: 25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13–16, 2018, proceedings, part III. Cham: Springer, Springer Nature. 2018. p. 435-447 doi: 10.1007/978-3-030-04182-3_38

Task and instance quadratic ordering for active online multitask learning

Abstract

Conference

Keywords

Access to Document

Other files and links

Cite this