POMDPs for assisting homeless shelters – Computational and deployment challenges

Amulya Yadav, Hau Chan, Albert Jiang, Eric Rice, Ece Kamar, Barbara Grosz, Milind Tambe

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

1 Citation (Scopus)

Abstract

This paper looks at challenges faced during the ongoing deployment of HEALER, a POMDP based software agent that recommends sequential intervention plans for use by homeless shelters, who organize these interventions to raise awareness about HIV among homeless youth. HEALER’s sequential plans (built using knowledge of social networks of homeless youth) choose intervention participants strategically to maximize influence spread, while reasoning about uncertainties in the network. In order to compute its plans, HEALER (i) casts this influence maximization problem as a POMDP and solves it using a novel planner which scales up to previously unsolvable real-world sizes; (ii) and constructs social networks of homeless youth at low cost, using a Facebook application. HEALER is currently being deployed in the real world in collaboration with a homeless shelter. Initial feedback from the shelter officials has been positive but they were surprised by the solutions generated by HEALER as these solutions are very counter-intuitive. Therefore, there is a need to justify HEALER’s solutions in a way that mirrors the officials’ intuition. In this paper, we report on progress made towards HEALER’s deployment and detail first steps taken to tackle the issue of explaining HEALER’s solutions.

Original languageEnglish (US)
Title of host publicationAutonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers
EditorsNardine Osman, Carles Sierra
PublisherSpringer Verlag
Pages67-87
Number of pages21
ISBN (Print)9783319468396
DOIs
StatePublished - Jan 1 2016
EventInternational Foundation for Autonomous Agents and Multiagent Systems, IFAAMAS 2016 - Singapore, Singapore
Duration: May 9 2016May 10 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10003 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceInternational Foundation for Autonomous Agents and Multiagent Systems, IFAAMAS 2016
CountrySingapore
CitySingapore
Period5/9/165/10/16

Fingerprint

Partially Observable Markov Decision Process
Social Networks
Software Agents
Scale-up
Justify
Software agents
Intuitive
Mirror
Reasoning
Choose
Maximise
Uncertainty
Feedback
Costs
Influence

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Yadav, A., Chan, H., Jiang, A., Rice, E., Kamar, E., Grosz, B., & Tambe, M. (2016). POMDPs for assisting homeless shelters – Computational and deployment challenges. In N. Osman, & C. Sierra (Eds.), Autonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers (pp. 67-87). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10003 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-319-46840-2_5
Yadav, Amulya ; Chan, Hau ; Jiang, Albert ; Rice, Eric ; Kamar, Ece ; Grosz, Barbara ; Tambe, Milind. / POMDPs for assisting homeless shelters – Computational and deployment challenges. Autonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers. editor / Nardine Osman ; Carles Sierra. Springer Verlag, 2016. pp. 67-87 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{b9d34a3bd1604fa4b0eaca748184f1bc,
title = "POMDPs for assisting homeless shelters – Computational and deployment challenges",
abstract = "This paper looks at challenges faced during the ongoing deployment of HEALER, a POMDP based software agent that recommends sequential intervention plans for use by homeless shelters, who organize these interventions to raise awareness about HIV among homeless youth. HEALER’s sequential plans (built using knowledge of social networks of homeless youth) choose intervention participants strategically to maximize influence spread, while reasoning about uncertainties in the network. In order to compute its plans, HEALER (i) casts this influence maximization problem as a POMDP and solves it using a novel planner which scales up to previously unsolvable real-world sizes; (ii) and constructs social networks of homeless youth at low cost, using a Facebook application. HEALER is currently being deployed in the real world in collaboration with a homeless shelter. Initial feedback from the shelter officials has been positive but they were surprised by the solutions generated by HEALER as these solutions are very counter-intuitive. Therefore, there is a need to justify HEALER’s solutions in a way that mirrors the officials’ intuition. In this paper, we report on progress made towards HEALER’s deployment and detail first steps taken to tackle the issue of explaining HEALER’s solutions.",
author = "Amulya Yadav and Hau Chan and Albert Jiang and Eric Rice and Ece Kamar and Barbara Grosz and Milind Tambe",
year = "2016",
month = "1",
day = "1",
doi = "10.1007/978-3-319-46840-2_5",
language = "English (US)",
isbn = "9783319468396",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "67--87",
editor = "Nardine Osman and Carles Sierra",
booktitle = "Autonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers",
address = "Germany",

}

Yadav, A, Chan, H, Jiang, A, Rice, E, Kamar, E, Grosz, B & Tambe, M 2016, POMDPs for assisting homeless shelters – Computational and deployment challenges. in N Osman & C Sierra (eds), Autonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10003 LNAI, Springer Verlag, pp. 67-87, International Foundation for Autonomous Agents and Multiagent Systems, IFAAMAS 2016, Singapore, Singapore, 5/9/16. https://doi.org/10.1007/978-3-319-46840-2_5

POMDPs for assisting homeless shelters – Computational and deployment challenges. / Yadav, Amulya; Chan, Hau; Jiang, Albert; Rice, Eric; Kamar, Ece; Grosz, Barbara; Tambe, Milind.

Autonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers. ed. / Nardine Osman; Carles Sierra. Springer Verlag, 2016. p. 67-87 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10003 LNAI).

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

TY - GEN

T1 - POMDPs for assisting homeless shelters – Computational and deployment challenges

AU - Yadav, Amulya

AU - Chan, Hau

AU - Jiang, Albert

AU - Rice, Eric

AU - Kamar, Ece

AU - Grosz, Barbara

AU - Tambe, Milind

PY - 2016/1/1

Y1 - 2016/1/1

N2 - This paper looks at challenges faced during the ongoing deployment of HEALER, a POMDP based software agent that recommends sequential intervention plans for use by homeless shelters, who organize these interventions to raise awareness about HIV among homeless youth. HEALER’s sequential plans (built using knowledge of social networks of homeless youth) choose intervention participants strategically to maximize influence spread, while reasoning about uncertainties in the network. In order to compute its plans, HEALER (i) casts this influence maximization problem as a POMDP and solves it using a novel planner which scales up to previously unsolvable real-world sizes; (ii) and constructs social networks of homeless youth at low cost, using a Facebook application. HEALER is currently being deployed in the real world in collaboration with a homeless shelter. Initial feedback from the shelter officials has been positive but they were surprised by the solutions generated by HEALER as these solutions are very counter-intuitive. Therefore, there is a need to justify HEALER’s solutions in a way that mirrors the officials’ intuition. In this paper, we report on progress made towards HEALER’s deployment and detail first steps taken to tackle the issue of explaining HEALER’s solutions.

AB - This paper looks at challenges faced during the ongoing deployment of HEALER, a POMDP based software agent that recommends sequential intervention plans for use by homeless shelters, who organize these interventions to raise awareness about HIV among homeless youth. HEALER’s sequential plans (built using knowledge of social networks of homeless youth) choose intervention participants strategically to maximize influence spread, while reasoning about uncertainties in the network. In order to compute its plans, HEALER (i) casts this influence maximization problem as a POMDP and solves it using a novel planner which scales up to previously unsolvable real-world sizes; (ii) and constructs social networks of homeless youth at low cost, using a Facebook application. HEALER is currently being deployed in the real world in collaboration with a homeless shelter. Initial feedback from the shelter officials has been positive but they were surprised by the solutions generated by HEALER as these solutions are very counter-intuitive. Therefore, there is a need to justify HEALER’s solutions in a way that mirrors the officials’ intuition. In this paper, we report on progress made towards HEALER’s deployment and detail first steps taken to tackle the issue of explaining HEALER’s solutions.

UR - http://www.scopus.com/inward/record.url?scp=84989868492&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84989868492&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-46840-2_5

DO - 10.1007/978-3-319-46840-2_5

M3 - Conference contribution

AN - SCOPUS:84989868492

SN - 9783319468396

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 67

EP - 87

BT - Autonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers

A2 - Osman, Nardine

A2 - Sierra, Carles

PB - Springer Verlag

ER -

Yadav A, Chan H, Jiang A, Rice E, Kamar E, Grosz B et al. POMDPs for assisting homeless shelters – Computational and deployment challenges. In Osman N, Sierra C, editors, Autonomous Agents and Multiagent Systems, IFAAMAS 2016 Workshops, Best Papers, Revised Selected Papers. Springer Verlag. 2016. p. 67-87. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-46840-2_5