Synthesizing Action Sequences for Modifying Model Decisions


  • Goutham Ramakrishnan University of Wisconsin-Madison
  • Yun Chan Lee University of Wisconsin-Madison
  • Aws Albarghouthi University of Wisconsin-Madison



When a model makes a consequential decision, e.g., denying someone a loan, it needs to additionally generate actionable, realistic feedback on what the person can do to favorably change the decision. We cast this problem through the lens of program synthesis, in which our goal is to synthesize an optimal (realistically cheapest or simplest) sequence of actions that if a person executes successfully can change their classification. We present a novel and general approach that combines search-based program synthesis and test-time adversarial attacks to construct action sequences over a domain-specific set of actions. We demonstrate the effectiveness of our approach on a number of deep neural networks.




How to Cite

Ramakrishnan, G., Lee, Y. C., & Albarghouthi, A. (2020). Synthesizing Action Sequences for Modifying Model Decisions. Proceedings of the AAAI Conference on Artificial Intelligence, 34(04), 5462-5469.



AAAI Technical Track: Machine Learning