Learning fairness under constraints

A decentralized resource allocation game

Qinyun Zhu, Jae C Oh

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

3 Citations (Scopus)

Abstract

We study multi-type resource allocation in multiagent system, where some constraints are enforced upon resource providers and users. These constraints are limitations of resource types and connection availabilities, which may make the collaboration between agents infeasible. We discuss the notion of distributed resource fairness under these constraints. Then we propose a game theory and reinforcement learning based solution for collaborative resource allocation, so that resources are assigned to users fairly and tasks are assigned to resource agents efficiently. We utilize data from Google data center as our input to simulations. Results show that our learning approach outperforms a greedy and random explorations in terms of resource utilization and fairness.

Original languageEnglish (US)
Title of host publicationProceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages214-221
Number of pages8
ISBN (Electronic)9781509061662
DOIs
StatePublished - Jan 31 2017
Event15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016 - Anaheim, United States
Duration: Dec 18 2016Dec 20 2016

Other

Other15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016
CountryUnited States
CityAnaheim
Period12/18/1612/20/16

Fingerprint

Resource allocation
Game theory
Reinforcement learning
Multi agent systems
Availability

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Computer Science Applications

Cite this

Zhu, Q., & Oh, J. C. (2017). Learning fairness under constraints: A decentralized resource allocation game. In Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016 (pp. 214-221). [7838147] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICMLA.2016.122

Learning fairness under constraints : A decentralized resource allocation game. / Zhu, Qinyun; Oh, Jae C.

Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016. Institute of Electrical and Electronics Engineers Inc., 2017. p. 214-221 7838147.

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

Zhu, Q & Oh, JC 2017, Learning fairness under constraints: A decentralized resource allocation game. in Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016., 7838147, Institute of Electrical and Electronics Engineers Inc., pp. 214-221, 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016, Anaheim, United States, 12/18/16. https://doi.org/10.1109/ICMLA.2016.122
Zhu Q, Oh JC. Learning fairness under constraints: A decentralized resource allocation game. In Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016. Institute of Electrical and Electronics Engineers Inc. 2017. p. 214-221. 7838147 https://doi.org/10.1109/ICMLA.2016.122
Zhu, Qinyun ; Oh, Jae C. / Learning fairness under constraints : A decentralized resource allocation game. Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 214-221
@inproceedings{ae63a7e468ea4a19a9941c1729ada9e9,
title = "Learning fairness under constraints: A decentralized resource allocation game",
abstract = "We study multi-type resource allocation in multiagent system, where some constraints are enforced upon resource providers and users. These constraints are limitations of resource types and connection availabilities, which may make the collaboration between agents infeasible. We discuss the notion of distributed resource fairness under these constraints. Then we propose a game theory and reinforcement learning based solution for collaborative resource allocation, so that resources are assigned to users fairly and tasks are assigned to resource agents efficiently. We utilize data from Google data center as our input to simulations. Results show that our learning approach outperforms a greedy and random explorations in terms of resource utilization and fairness.",
author = "Qinyun Zhu and Oh, {Jae C}",
year = "2017",
month = "1",
day = "31",
doi = "10.1109/ICMLA.2016.122",
language = "English (US)",
pages = "214--221",
booktitle = "Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Learning fairness under constraints

T2 - A decentralized resource allocation game

AU - Zhu, Qinyun

AU - Oh, Jae C

PY - 2017/1/31

Y1 - 2017/1/31

N2 - We study multi-type resource allocation in multiagent system, where some constraints are enforced upon resource providers and users. These constraints are limitations of resource types and connection availabilities, which may make the collaboration between agents infeasible. We discuss the notion of distributed resource fairness under these constraints. Then we propose a game theory and reinforcement learning based solution for collaborative resource allocation, so that resources are assigned to users fairly and tasks are assigned to resource agents efficiently. We utilize data from Google data center as our input to simulations. Results show that our learning approach outperforms a greedy and random explorations in terms of resource utilization and fairness.

AB - We study multi-type resource allocation in multiagent system, where some constraints are enforced upon resource providers and users. These constraints are limitations of resource types and connection availabilities, which may make the collaboration between agents infeasible. We discuss the notion of distributed resource fairness under these constraints. Then we propose a game theory and reinforcement learning based solution for collaborative resource allocation, so that resources are assigned to users fairly and tasks are assigned to resource agents efficiently. We utilize data from Google data center as our input to simulations. Results show that our learning approach outperforms a greedy and random explorations in terms of resource utilization and fairness.

UR - http://www.scopus.com/inward/record.url?scp=85015420051&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85015420051&partnerID=8YFLogxK

U2 - 10.1109/ICMLA.2016.122

DO - 10.1109/ICMLA.2016.122

M3 - Conference contribution

SP - 214

EP - 221

BT - Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016

PB - Institute of Electrical and Electronics Engineers Inc.

ER -