Abstract

The PART-WHOLE relationship routinely finds itself in many disciplines, ranging from collaborative teams, crowdsourcing, autonomous systems to networked systems. From the algorithmic perspective, the existing work has primarily focused on predicting the outcomes of the whole and parts, by either separate models or linear joint models, which assume the outcome of the parts has a linear and independent effect on the outcome of the whole. In this paper, we propose a joint predictive method named PAROLE to simultaneously and mutually predict the part and whole outcomes. The proposed method offers two distinct advantages over the existing work. First (Model Generality), we formulate joint PART-WHOLE outcome prediction as a generic optimization problem, which is able to encode a variety of complex relationships between the outcome of the whole and parts, beyond the linear independence assumption. Second (Algorithm Efficacy), we propose an effective and efficient block coordinate descent algorithm, which is able to find the coordinate-wise optimum with a linear complexity in both time and space. Extensive empirical evaluations on real-world datasets demonstrate that the proposed PAROLE (1) leads to consistent prediction performance improvement by modeling the non-linear part-whole relationship as well as part-part interdependency, and (2) scales linearly in terms of the size of the training dataset.

Original languageEnglish (US)
Title of host publicationKDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
PublisherAssociation for Computing Machinery
Pages295-304
Number of pages10
VolumePart F129685
ISBN (Electronic)9781450348874
DOIs
StatePublished - Aug 13 2017
Event23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2017 - Halifax, Canada
Duration: Aug 13 2017Aug 17 2017

Other

Other23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2017
CountryCanada
CityHalifax
Period8/13/178/17/17

Keywords

  • Joint predictive model
  • Part-whole relationship

ASJC Scopus subject areas

  • Software
  • Information Systems

Cite this

Li, L., Tong, H., Wang, Y., Shi, C., Cao, N., & Buchler, N. (2017). Is the whole greater than the sum of its parts? In KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (Vol. Part F129685, pp. 295-304). Association for Computing Machinery. https://doi.org/10.1145/3097983.3098006

Is the whole greater than the sum of its parts? / Li, Liangyue; Tong, Hanghang; Wang, Yong; Shi, Conglei; Cao, Nan; Buchler, Norbou.

KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. Part F129685 Association for Computing Machinery, 2017. p. 295-304.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Li, L, Tong, H, Wang, Y, Shi, C, Cao, N & Buchler, N 2017, Is the whole greater than the sum of its parts? in KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. vol. Part F129685, Association for Computing Machinery, pp. 295-304, 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2017, Halifax, Canada, 8/13/17. https://doi.org/10.1145/3097983.3098006
Li L, Tong H, Wang Y, Shi C, Cao N, Buchler N. Is the whole greater than the sum of its parts? In KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. Part F129685. Association for Computing Machinery. 2017. p. 295-304 https://doi.org/10.1145/3097983.3098006
Li, Liangyue ; Tong, Hanghang ; Wang, Yong ; Shi, Conglei ; Cao, Nan ; Buchler, Norbou. / Is the whole greater than the sum of its parts?. KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. Part F129685 Association for Computing Machinery, 2017. pp. 295-304
@inproceedings{8dbf730f31534a25bada6774ec543ca1,
title = "Is the whole greater than the sum of its parts?",
abstract = "The PART-WHOLE relationship routinely finds itself in many disciplines, ranging from collaborative teams, crowdsourcing, autonomous systems to networked systems. From the algorithmic perspective, the existing work has primarily focused on predicting the outcomes of the whole and parts, by either separate models or linear joint models, which assume the outcome of the parts has a linear and independent effect on the outcome of the whole. In this paper, we propose a joint predictive method named PAROLE to simultaneously and mutually predict the part and whole outcomes. The proposed method offers two distinct advantages over the existing work. First (Model Generality), we formulate joint PART-WHOLE outcome prediction as a generic optimization problem, which is able to encode a variety of complex relationships between the outcome of the whole and parts, beyond the linear independence assumption. Second (Algorithm Efficacy), we propose an effective and efficient block coordinate descent algorithm, which is able to find the coordinate-wise optimum with a linear complexity in both time and space. Extensive empirical evaluations on real-world datasets demonstrate that the proposed PAROLE (1) leads to consistent prediction performance improvement by modeling the non-linear part-whole relationship as well as part-part interdependency, and (2) scales linearly in terms of the size of the training dataset.",
keywords = "Joint predictive model, Part-whole relationship",
author = "Liangyue Li and Hanghang Tong and Yong Wang and Conglei Shi and Nan Cao and Norbou Buchler",
year = "2017",
month = "8",
day = "13",
doi = "10.1145/3097983.3098006",
language = "English (US)",
volume = "Part F129685",
pages = "295--304",
booktitle = "KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining",
publisher = "Association for Computing Machinery",

}

TY - GEN

T1 - Is the whole greater than the sum of its parts?

AU - Li, Liangyue

AU - Tong, Hanghang

AU - Wang, Yong

AU - Shi, Conglei

AU - Cao, Nan

AU - Buchler, Norbou

PY - 2017/8/13

Y1 - 2017/8/13

N2 - The PART-WHOLE relationship routinely finds itself in many disciplines, ranging from collaborative teams, crowdsourcing, autonomous systems to networked systems. From the algorithmic perspective, the existing work has primarily focused on predicting the outcomes of the whole and parts, by either separate models or linear joint models, which assume the outcome of the parts has a linear and independent effect on the outcome of the whole. In this paper, we propose a joint predictive method named PAROLE to simultaneously and mutually predict the part and whole outcomes. The proposed method offers two distinct advantages over the existing work. First (Model Generality), we formulate joint PART-WHOLE outcome prediction as a generic optimization problem, which is able to encode a variety of complex relationships between the outcome of the whole and parts, beyond the linear independence assumption. Second (Algorithm Efficacy), we propose an effective and efficient block coordinate descent algorithm, which is able to find the coordinate-wise optimum with a linear complexity in both time and space. Extensive empirical evaluations on real-world datasets demonstrate that the proposed PAROLE (1) leads to consistent prediction performance improvement by modeling the non-linear part-whole relationship as well as part-part interdependency, and (2) scales linearly in terms of the size of the training dataset.

AB - The PART-WHOLE relationship routinely finds itself in many disciplines, ranging from collaborative teams, crowdsourcing, autonomous systems to networked systems. From the algorithmic perspective, the existing work has primarily focused on predicting the outcomes of the whole and parts, by either separate models or linear joint models, which assume the outcome of the parts has a linear and independent effect on the outcome of the whole. In this paper, we propose a joint predictive method named PAROLE to simultaneously and mutually predict the part and whole outcomes. The proposed method offers two distinct advantages over the existing work. First (Model Generality), we formulate joint PART-WHOLE outcome prediction as a generic optimization problem, which is able to encode a variety of complex relationships between the outcome of the whole and parts, beyond the linear independence assumption. Second (Algorithm Efficacy), we propose an effective and efficient block coordinate descent algorithm, which is able to find the coordinate-wise optimum with a linear complexity in both time and space. Extensive empirical evaluations on real-world datasets demonstrate that the proposed PAROLE (1) leads to consistent prediction performance improvement by modeling the non-linear part-whole relationship as well as part-part interdependency, and (2) scales linearly in terms of the size of the training dataset.

KW - Joint predictive model

KW - Part-whole relationship

UR - http://www.scopus.com/inward/record.url?scp=85029044613&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029044613&partnerID=8YFLogxK

U2 - 10.1145/3097983.3098006

DO - 10.1145/3097983.3098006

M3 - Conference contribution

AN - SCOPUS:85029044613

VL - Part F129685

SP - 295

EP - 304

BT - KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

PB - Association for Computing Machinery

ER -