Using Reinforcement Learning to Optimize Responses in Care Processes: A Case Study on Aggression Incidents

Bart J. Verhoef; Xixi Lu

doi:10.1007/978-3-031-50974-2_5

Using Reinforcement Learning to Optimize Responses in Care Processes: A Case Study on Aggression Incidents

Bart J. Verhoef, Xixi Lu^*

^*Corresponding author for this work

Utrecht University

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

Abstract

Previous studies have used prescriptive process monitoring to find actionable policies in business processes and conducted case studies in similar domains, such as the loan application process and the traffic fine process. However, care processes tend to be more dynamic and complex. For example, at any stage of a care process, a multitude of actions is possible. In this paper, we follow the reinforcement approach and train a Markov decision process using event data from a care process. The goal was to find optimal policies for staff members when clients are displaying any type of aggressive behavior. We used the reinforcement learning algorithms Q-learning and SARSA to find optimal policies. Results showed that the policies derived from these algorithms are similar to the most frequent actions currently used but provide the staff members with a few more options in certain situations.

Original language	English
Title of host publication	Business Process Management Workshops
Subtitle of host publication	BPM 2023 International Workshops, Utrecht, The Netherlands, September 11–15, 2023, Revised Selected Papers
Editors	Jochen De Weerdt, Luise Pufahl
Publisher	Springer
Pages	57-69
Number of pages	13
ISBN (Print)	9783031509735
DOIs	https://doi.org/10.1007/978-3-031-50974-2_5
Publication status	Published - 11 Jan 2024
Event	International Workshops held at the 21st International Conference on Business Process Management, BPM 2023 - Utrecht, Netherlands Duration: 11 Sept 2023 → 15 Sept 2023

Publication series

Name	Lecture Notes in Business Information Processing
Volume	492 LNBIP
ISSN (Print)	1865-1348
ISSN (Electronic)	1865-1356

Conference

Conference	International Workshops held at the 21st International Conference on Business Process Management, BPM 2023
Country/Territory	Netherlands
City	Utrecht
Period	11/09/23 → 15/09/23

Bibliographical note

Publisher Copyright:
© 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Keywords

Markov decision process
prescriptive process mining
process mining
process optimization
reinforcement learning

Access to Document

10.1007/978-3-031-50974-2_5

Cite this

Verhoef, B. J., & Lu, X. (2024). Using Reinforcement Learning to Optimize Responses in Care Processes: A Case Study on Aggression Incidents. In J. De Weerdt, & L. Pufahl (Eds.), Business Process Management Workshops : BPM 2023 International Workshops, Utrecht, The Netherlands, September 11–15, 2023, Revised Selected Papers (pp. 57-69). (Lecture Notes in Business Information Processing; Vol. 492 LNBIP). Springer. https://doi.org/10.1007/978-3-031-50974-2_5

Verhoef, Bart J. ; Lu, Xixi. / Using Reinforcement Learning to Optimize Responses in Care Processes : A Case Study on Aggression Incidents. Business Process Management Workshops : BPM 2023 International Workshops, Utrecht, The Netherlands, September 11–15, 2023, Revised Selected Papers. editor / Jochen De Weerdt ; Luise Pufahl. Springer, 2024. pp. 57-69 (Lecture Notes in Business Information Processing).

@inproceedings{70c47f7548ea4ffc93c746e34007df94,

title = "Using Reinforcement Learning to Optimize Responses in Care Processes: A Case Study on Aggression Incidents",

abstract = "Previous studies have used prescriptive process monitoring to find actionable policies in business processes and conducted case studies in similar domains, such as the loan application process and the traffic fine process. However, care processes tend to be more dynamic and complex. For example, at any stage of a care process, a multitude of actions is possible. In this paper, we follow the reinforcement approach and train a Markov decision process using event data from a care process. The goal was to find optimal policies for staff members when clients are displaying any type of aggressive behavior. We used the reinforcement learning algorithms Q-learning and SARSA to find optimal policies. Results showed that the policies derived from these algorithms are similar to the most frequent actions currently used but provide the staff members with a few more options in certain situations.",

keywords = "Markov decision process, prescriptive process mining, process mining, process optimization, reinforcement learning",

author = "Verhoef, {Bart J.} and Xixi Lu",

note = "Publisher Copyright: {\textcopyright} 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.; International Workshops held at the 21st International Conference on Business Process Management, BPM 2023 ; Conference date: 11-09-2023 Through 15-09-2023",

year = "2024",

month = jan,

day = "11",

doi = "10.1007/978-3-031-50974-2_5",

language = "English",

isbn = "9783031509735",

series = "Lecture Notes in Business Information Processing",

publisher = "Springer",

pages = "57--69",

editor = "{De Weerdt}, Jochen and Luise Pufahl",

booktitle = "Business Process Management Workshops",

}

Verhoef, BJ & Lu, X 2024, Using Reinforcement Learning to Optimize Responses in Care Processes: A Case Study on Aggression Incidents. in J De Weerdt & L Pufahl (eds), Business Process Management Workshops : BPM 2023 International Workshops, Utrecht, The Netherlands, September 11–15, 2023, Revised Selected Papers. Lecture Notes in Business Information Processing, vol. 492 LNBIP, Springer, pp. 57-69, International Workshops held at the 21st International Conference on Business Process Management, BPM 2023, Utrecht, Netherlands, 11/09/23. https://doi.org/10.1007/978-3-031-50974-2_5

Using Reinforcement Learning to Optimize Responses in Care Processes: A Case Study on Aggression Incidents. / Verhoef, Bart J.; Lu, Xixi.
Business Process Management Workshops : BPM 2023 International Workshops, Utrecht, The Netherlands, September 11–15, 2023, Revised Selected Papers. ed. / Jochen De Weerdt; Luise Pufahl. Springer, 2024. p. 57-69 (Lecture Notes in Business Information Processing; Vol. 492 LNBIP).

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

TY - GEN

T1 - Using Reinforcement Learning to Optimize Responses in Care Processes

T2 - International Workshops held at the 21st International Conference on Business Process Management, BPM 2023

AU - Verhoef, Bart J.

AU - Lu, Xixi

PY - 2024/1/11

Y1 - 2024/1/11

N2 - Previous studies have used prescriptive process monitoring to find actionable policies in business processes and conducted case studies in similar domains, such as the loan application process and the traffic fine process. However, care processes tend to be more dynamic and complex. For example, at any stage of a care process, a multitude of actions is possible. In this paper, we follow the reinforcement approach and train a Markov decision process using event data from a care process. The goal was to find optimal policies for staff members when clients are displaying any type of aggressive behavior. We used the reinforcement learning algorithms Q-learning and SARSA to find optimal policies. Results showed that the policies derived from these algorithms are similar to the most frequent actions currently used but provide the staff members with a few more options in certain situations.

AB - Previous studies have used prescriptive process monitoring to find actionable policies in business processes and conducted case studies in similar domains, such as the loan application process and the traffic fine process. However, care processes tend to be more dynamic and complex. For example, at any stage of a care process, a multitude of actions is possible. In this paper, we follow the reinforcement approach and train a Markov decision process using event data from a care process. The goal was to find optimal policies for staff members when clients are displaying any type of aggressive behavior. We used the reinforcement learning algorithms Q-learning and SARSA to find optimal policies. Results showed that the policies derived from these algorithms are similar to the most frequent actions currently used but provide the staff members with a few more options in certain situations.

KW - Markov decision process

KW - prescriptive process mining

KW - process mining

KW - process optimization

KW - reinforcement learning

UR - http://www.scopus.com/inward/record.url?scp=85182587406&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-50974-2_5

DO - 10.1007/978-3-031-50974-2_5

M3 - Conference contribution

AN - SCOPUS:85182587406

SN - 9783031509735

T3 - Lecture Notes in Business Information Processing

SP - 57

EP - 69

BT - Business Process Management Workshops

A2 - De Weerdt, Jochen

A2 - Pufahl, Luise

PB - Springer

Y2 - 11 September 2023 through 15 September 2023

ER -

Verhoef BJ, Lu X. Using Reinforcement Learning to Optimize Responses in Care Processes: A Case Study on Aggression Incidents. In De Weerdt J, Pufahl L, editors, Business Process Management Workshops : BPM 2023 International Workshops, Utrecht, The Netherlands, September 11–15, 2023, Revised Selected Papers. Springer. 2024. p. 57-69. (Lecture Notes in Business Information Processing). doi: 10.1007/978-3-031-50974-2_5