Grant: Discovery Grants Program - Individual
University of Waterloo
$47,000.00
Advances in Algorithmic Recourse: Multi-Agent Dynamics, Stakeholder Ethics, and Evaluative Standards
EF 2024-25
Résumé
In today's digital age, machines are not just lifting heavy objects or doing repetitive tasks - they are also making decisions that significantly impact people's lives. Consider someone being refused a loan, or denied a specific medical treatment, not due to a human's evaluation finding concerns, but because a machine determined it - without providing a clear explanation for that choice. Artificial Intelligence (AI) is swiftly making such determinations in domains like court bail decisions, financial loans, and even recruitment processes and health advice. Yet, how can individuals place trust in these machines if they remain opaque in their functioning?
To enhance trust in AI among Canadians, the Government of Canada introduced a Directive on Automated Decision Making promoting transparent AI decisions through Explainable AI (XAI). Over the past 5 years, Dr. Karimi's team has crafted solutions to explain a variety of AI systems, offered as sets of actionable recommendations termed algorithmic recourse. These are anchored in rigorous mathematics and engineering, addressing issues such as loan denials. Dr. Karimi's seminal work in this field has been showcased at major international forums, marking him as a leader in responsible AI and XAI research. Based on this groundwork, the proposed program's aims are:
1. Extend recourse methods to evolving environments, and for non-learning AI systems, like matching and scheduling.
2. Craft algorithms balancing individual recourse with group-focused criteria like fairness, robustness, and privacy.
3. Establish metrics to assess and validate the applicability domain of algorithmic recourse methods.
Dr. Karimi's program serves dual purposes: (i) it aims to contribute to both the research community and fortifies Canada's position in responsible AI advancements; (ii) it aims to train the next wave of skilled XAI experts. Through his program, students will delve into and expand upon foundational XAI research, setting them on a trajectory for flourishing careers in a rapidly evolving sector that holds immense promise for Canada.
Beyond simply adhering to the current regulatory standards, Dr. Karimi envisions a future where AI systems seamlessly collaborate with and augment human potential. Analogous to how we trust everyday appliances, like dishwashers, to execute their functions reliably and without constant attention, Dr. Karimi foresees AI systems taking on complex cognitive responsibilities, running smoothly without the need for perpetual human intervention. Just as engineered systems are expected to operate seamlessly within their specified domain and handle failures gracefully, so too should human-AI systems work within their designated parameters, yet also have the ability to offer intuitive, easily understandable guidance when confronted with unpredictabilities or errors. With a focus on the latter, Dr. Karimi's research program is strategically positioned to achieve the broader long-term vision.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2024-05160
Name nom
Karimi, Amir-Hossein
Department département
Computer Science
Organization id
33
Institution établissement
University of Waterloo
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
47000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1206
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Engineering
Area of application fr
Génie
Research subject code
2800
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Artificial intelligence (computer vision, use2603)
Research subject fr
Intelligence artificielle (vision artificielle, utiliser 2603)
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Advances in Algorithmic Recourse: Multi-Agent Dynamics, Stakeholder Ethics, and Evaluative Standards
Keywords
artificial intelligence; machine learning; explainable ai; interpretable ml; causal inference; algorithmic recourse; counterfactual explanations; program synthesis; human-machine teaming; human-machine collaboration
Application summary
In today's digital age, machines are not just lifting heavy objects or doing repetitive tasks - they are also making decisions that significantly impact people's lives. Consider someone being refused a loan, or denied a specific medical treatment, not due to a human's evaluation finding concerns, but because a machine determined it - without providing a clear explanation for that choice. Artificial Intelligence (AI) is swiftly making such determinations in domains like court bail decisions, financial loans, and even recruitment processes and health advice. Yet, how can individuals place trust in these machines if they remain opaque in their functioning?
To enhance trust in AI among Canadians, the Government of Canada introduced a Directive on Automated Decision Making promoting transparent AI decisions through Explainable AI (XAI). Over the past 5 years, Dr. Karimi's team has crafted solutions to explain a variety of AI systems, offered as sets of actionable recommendations termed algorithmic recourse. These are anchored in rigorous mathematics and engineering, addressing issues such as loan denials. Dr. Karimi's seminal work in this field has been showcased at major international forums, marking him as a leader in responsible AI and XAI research. Based on this groundwork, the proposed program's aims are:
1. Extend recourse methods to evolving environments, and for non-learning AI systems, like matching and scheduling.
2. Craft algorithms balancing individual recourse with group-focused criteria like fairness, robustness, and privacy.
3. Establish metrics to assess and validate the applicability domain of algorithmic recourse methods.
Dr. Karimi's program serves dual purposes: (i) it aims to contribute to both the research community and fortifies Canada's position in responsible AI advancements; (ii) it aims to train the next wave of skilled XAI experts. Through his program, students will delve into and expand upon foundational XAI research, setting them on a trajectory for flourishing careers in a rapidly evolving sector that holds immense promise for Canada.
Beyond simply adhering to the current regulatory standards, Dr. Karimi envisions a future where AI systems seamlessly collaborate with and augment human potential. Analogous to how we trust everyday appliances, like dishwashers, to execute their functions reliably and without constant attention, Dr. Karimi foresees AI systems taking on complex cognitive responsibilities, running smoothly without the need for perpetual human intervention. Just as engineered systems are expected to operate seamlessly within their specified domain and handle failures gracefully, so too should human-AI systems work within their designated parameters, yet also have the ability to offer intuitive, easily understandable guidance when confronted with unpredictabilities or errors. With a focus on the latter, Dr. Karimi's research program is strategically positioned to achieve the broader long-term vision.