Sriram Gopalakrishnan, Ph.D.

Sriram Gopalakrishnan, Ph.D.

AI Research Lead at J.P. Morgan Chase | Sequential-Decision Making Problems, Human-AI Interaction, & Generative AI

SafeAR: Towards Safer Algorithmic Recourse by Risk-Aware Policies

Published in Proceedings of the AAAI Conference on Artificial Intelligence (AAAI-24), 2024

Recommended citation: Wu, H., Sharma, S., Patra, S., & Gopalakrishnan, S. (2024). SafeAR: Towards Safer Algorithmic Recourse by Risk-Aware Policies. In Proceedings of the AAAI Conference on Artificial Intelligence. https://ojs.aaai.org/index.php/AAAI/article/view/30206

🎙️ Oral Presentation (Top 2% of accepted papers)

Abstract

Algorithmic recourse provides actionable recommendations for individuals seeking to reverse unfavourable automated decisions. However, standard methods assume deterministic execution. We introduce SafeAR, a risk-aware sequential recourse framework that accounts for stochastic execution and minimizes downside risk via Conditional Value-at-Risk (CVaR) objectives.

View Publication

Recommended citation: Wu, H., Sharma, S., Patra, S., & Gopalakrishnan, S. (2024). SafeAR: Towards Safer Algorithmic Recourse by Risk-Aware Policies. In Proceedings of the AAAI Conference on Artificial Intelligence.