Search-based fairness testing for regression-based machine learning systems

Journal article


Perera, Anjana, Aleti, Aldeida, Tantithamthavorn, Chakkrit, Jiarpakdee, Jirayus, Turhan, Burak, Kuhn, Lisa and Walker, Katie. (2022). Search-based fairness testing for regression-based machine learning systems. Empirical Software Engineering. 27(3), p. Article 79. https://doi.org/10.1007/s10664-022-10116-7
AuthorsPerera, Anjana, Aleti, Aldeida, Tantithamthavorn, Chakkrit, Jiarpakdee, Jirayus, Turhan, Burak, Kuhn, Lisa and Walker, Katie
Abstract

Context
Machine learning (ML) software systems are permeating many aspects of our life, such as healthcare, transportation, banking, and recruitment. These systems are trained with data that is often biased, resulting in biased behaviour. To address this issue, fairness testing approaches have been proposed to test ML systems for fairness, which predominantly focus on assessing classification-based ML systems. These methods are not applicable to regression-based systems, for example, they do not quantify the magnitude of the disparity in predicted outcomes, which we identify as important in the context of regression-based ML systems.

Method:
We conduct this study as design science research. We identify the problem instance in the context of emergency department (ED) wait-time prediction. In this paper, we develop an effective and efficient fairness testing approach to evaluate the fairness of regression-based ML systems. We propose fairness degree, which is a new fairness measure for regression-based ML systems, and a novel search-based fairness testing (SBFT) approach for testing regression-based machine learning systems. We apply the proposed solutions to ED wait-time prediction software.

Results:
We experimentally evaluate the effectiveness and efficiency of the proposed approach with ML systems trained on real observational data from the healthcare domain. We demonstrate that SBFT significantly outperforms existing fairness testing approaches, with up to 111% and 190% increase in effectiveness and efficiency of SBFT compared to the best performing existing approaches.

Conclusion:
These findings indicate that our novel fairness measure and the new approach for fairness testing of regression-based ML systems can identify the degree of fairness in predictions, which can help software teams to make data-informed decisions about whether such software systems are ready to deploy. The scientific knowledge gained from our work can be phrased as a technological rule; to measure the fairness of the regression-based ML systems in the context of emergency department wait-time prediction use fairness degree and search-based techniques to approximate it.

Keywordsfairness testing; software testing; search-based software testing; software fairness; machine learning; bias
Year2022
JournalEmpirical Software Engineering
Journal citation27 (3), p. Article 79
PublisherSpringer
ISSN1382-3256
Digital Object Identifier (DOI)https://doi.org/10.1007/s10664-022-10116-7
Scopus EID2-s2.0-85127566250
Open accessPublished as ‘gold’ (paid) open access
Page range1-36
FunderAustralian Government
Medical Research Future Fund (MRFF), Australian Government
Monash Partners
Australian Research Council (ARC)
Publisher's version
License
File Access Level
Open
Output statusPublished
Publication dates
Online30 Mar 2022
Publication process dates
Accepted29 Dec 2021
Deposited04 Jul 2023
ARC Funded ResearchThis output has been funded, wholly or partially, under the Australian Research Council Act 2001
Grant IDDE200100941
Permalink -

https://acuresearchbank.acu.edu.au/item/8z379/search-based-fairness-testing-for-regression-based-machine-learning-systems

Download files


Publisher's version
  • 47
    total views
  • 27
    total downloads
  • 1
    views this month
  • 0
    downloads this month
These values are for the period from 19th October 2020, when this repository was created.

Export as

Related outputs

Vital signs monitoring in Australasian emergency departments : Development of a consensus statement from ACEM and CENA
Connell, Cliff, Craig, Simon, Crock, Carmel, Kuhn, Lisa, Morphet, Julia and Unwin, Maria. (2024). Vital signs monitoring in Australasian emergency departments : Development of a consensus statement from ACEM and CENA. Australasian Emergency Care. 27(3), pp. 207-217. https://doi.org/10.1016/j.auec.2024.04.001
Core competencies of emergency nurses for the armed conflict context : Experiences from the field
Mani, Zakaria A., Kuhn, Lisa and Plummer, Virginia. (2023). Core competencies of emergency nurses for the armed conflict context : Experiences from the field. International Nursing Review. pp. 1-10. https://doi.org/10.1111/inr.12902
Potential impact of a novel pathway for suspected myocardial infarction utilising a new high-sensitivity cardiac troponin I assay
Meek, Robert, Cullen, Louise, Lu, Zhong Xian, Nasis, Arthur, Kuhn, Lisa and Sorace, Laurence. (2022). Potential impact of a novel pathway for suspected myocardial infarction utilising a new high-sensitivity cardiac troponin I assay. Emergency Medicine Journal. 39(11), pp. 847-852. https://doi.org/10.1136/emermed-2020-210812
The relationship between social media usage by undergraduate nursing students and development of their professional identity : A correlational study
Alharbi, Muna, Kuhn, Lisa and Morphet, Julia. (2022). The relationship between social media usage by undergraduate nursing students and development of their professional identity : A correlational study. Nurse Education Today. 112, p. Article 105337. https://doi.org/10.1016/j.nedt.2022.105337
Predictors of radial to femoral artery access crossover during primary percutaneous coronary intervention for ST-elevation myocardial infarction
Dang, Denee, Kuhn, Lisa, Fooladi, Ensieh, Ky, Vivian, Cheung, Kevin, Rashid, Hashrul and Zaman, Sarah. (2022). Predictors of radial to femoral artery access crossover during primary percutaneous coronary intervention for ST-elevation myocardial infarction. Heart, Lung and Circulation. 31(7), pp. 985-992. https://doi.org/10.1016/j.hlc.2022.01.016
From little things, big things grow : An exploratory analysis of the national cost of peripheral intravenous catheter insertion in Australian adult emergency care
Morgan, Rachel, Callander, Emily, Cullen, Louise, Walker, Katie, Bumpstead, Suzanne, Hawkins, Tracey, Kuhn, Lisa and Egerton-Warburton, Diana. (2022). From little things, big things grow : An exploratory analysis of the national cost of peripheral intravenous catheter insertion in Australian adult emergency care. Emergency Medicine Australasia. 34(6), pp. 877-883. https://doi.org/10.1111/1742-6723.14009
Sex disparities in myocardial infarction : Biology or bias?
Stehli, Julia, Duffy, Stephen J., Burgess, Sonya, Kuhn, Lisa, Gulati, Martha, Chow, Clara and Zaman, Sarah. (2021). Sex disparities in myocardial infarction : Biology or bias? Heart, Lung and Circulation. 30(1), pp. 18-26. https://doi.org/10.1016/j.hlc.2020.06.025
The influence of nurse allocated triage category on the care of patients with sepsis in the emergency department : A retrospective review
Nevill, Alexandra, Kuhn, Lisa, Thompson, John and Morphet, Julia. (2021). The influence of nurse allocated triage category on the care of patients with sepsis in the emergency department : A retrospective review. Australasian Emergency Care. 24(2), pp. 121-126. https://doi.org/10.1016/j.auec.2020.09.002
The sourcing and use of high physical resemblance personal protective equipment to train healthcare workers, improve confidence and conserve medical-grade equipment
Bumpstead, S., Lim, Z. J., Kuhn, L., Flynn, D., Bakos, C.-L., Potter, E. and Egerton-Warburton, D.. (2021). The sourcing and use of high physical resemblance personal protective equipment to train healthcare workers, improve confidence and conserve medical-grade equipment. Journal of Hospital Infection. 112, pp. 104-107. https://doi.org/10.1016/j.jhin.2021.04.003
Nursing students' engagement with social media as an extracurricular activity : An integrative review
Alharbi, Muna, Kuhn, Lisa and Morphet, Julia. (2021). Nursing students' engagement with social media as an extracurricular activity : An integrative review. Journal of Clinical Nursing. 30(1-2), pp. 44-55. https://doi.org/10.1111/jocn.15503
The design and evaluation of a pre-procedure checklist specific to the cardiac catheterisation laboratory
Nicholson, Patricia, Kuhn, Lisa, Manias, Elizabeth and Sloman, Marie. (2021). The design and evaluation of a pre-procedure checklist specific to the cardiac catheterisation laboratory. Australian Critical Care. 34(4), pp. 350-357. https://doi.org/10.1016/j.aucc.2020.10.005
The nurses’ role in antimicrobial stewardship : A scoping review
van Huizen, Pheona, Kuhn, Lisa, Russo, Philip L. and Connell, Clifford J.. (2021). The nurses’ role in antimicrobial stewardship : A scoping review. International Journal of Nursing Studies. 113, p. Article 103772. https://doi.org/10.1016/j.ijnurstu.2020.103772
Nurse expertise : A critical resource in the COVID-19 pandemic response
Schwerdtl, Patricia Nayna, Connell, Clifford J., Lee, Susan, Plummer, Virginia, Russo, Philip L., Endacott, Ruth and Kuhn, Lisa. (2020). Nurse expertise : A critical resource in the COVID-19 pandemic response. Annals of Global Health. 86(1), pp. 1-5. https://doi.org/10.5334/aogh.2898
Recent trends in heroin and pharmaceutical opioid-related harms in Victoria, Australia up to 2018
Lam, Tina, Kuhn, Lisa, Hayman, Jane, Middleton, Melissa, Wilson, James, Scott, Debbie, Lubman, Dan, Smith, Karen and Nielsen, Suzanne. (2020). Recent trends in heroin and pharmaceutical opioid-related harms in Victoria, Australia up to 2018. Addiction. 115(2), pp. 261-269. https://doi.org/10.1111/add.14784
Common domains of core competencies for hospital health care providers in armed conflict zones : A systematic scoping review
Mani, Zakaria A., Kuhn, Lisa and Plummer, Virginia. (2020). Common domains of core competencies for hospital health care providers in armed conflict zones : A systematic scoping review. Prehospital and Disaster Medicine. 35(4), pp. 442-446. https://doi.org/10.1017/S1049023X20000503
A systematic review of opioid overdose interventions delivered within emergency departments
Chen, Yanjin, Wang, Yanbin, Nielsen, Suzanne, Kuhn, Lisa and Lam, Tina. (2020). A systematic review of opioid overdose interventions delivered within emergency departments. Drug and Alcohol Dependence. 213, p. Article 108009. https://doi.org/10.1016/j.drugalcdep.2020.108009
Physical health assessment and cardiometabolic monitoring practices across three adult mental health inpatient units – a retrospective cohort study
Howard, Rebekah, Kuhn, Lisa, Millar, Freyja and Street, Maryann. (2020). Physical health assessment and cardiometabolic monitoring practices across three adult mental health inpatient units – a retrospective cohort study. International Journal of Mental Health Nursing. 29(6), pp. 1144-1156. https://doi.org/10.1111/inm.12755
Undergraduate nursing students' adoption of the professional identity of nursing through social media use : A qualitative descriptive study
Alharbi, Muna, Kuhn, Lisa and Morphet, Julia. (2020). Undergraduate nursing students' adoption of the professional identity of nursing through social media use : A qualitative descriptive study. Nurse Education Today. 92, p. Article 104488. https://doi.org/10.1016/j.nedt.2020.104488
Effect of gender on evidence-based practice for Australian patients with acute coronary syndrome : A retrospective multi-site study
Kuhn, Lisa, Page, Karen, Street, Maryann, Rolley, John and Considine, Julie. (2017). Effect of gender on evidence-based practice for Australian patients with acute coronary syndrome : A retrospective multi-site study. Australasian Emergency Nursing Journal. 20(2), pp. 63-68. https://doi.org/10.1016/j.aenj.2017.02.002