Computer Science Thesis Oral
In Person - McWilliams Classroom, Gates Hillman 4303
HANRUI ZHANG , Ph.D. Candidate, Computer Science Department, Carnegie Mellon University
Designing and Analyzing Machine Learning Algorithms in the Presence of Strategic Behavior
Machine learning algorithms now play a major part in all kinds of decision-making scenarios. When the stakes are high, self-interested agents — about whom decisions are being made — are increasingly tempted to manipulate the machine learning algorithm, in order to better fulfill their own goals, which are generally different from the decision maker's. This highlights the importance of making machine learning algorithms robust against manipulation.
In this talk, I will focus on generalization (i.e., the bridge between training and testing) in strategic classification: Traditional wisdom suggests that a classifier trained on historical observations (i.e., the training set) usually also works well on future data points to be classified (i.e., the test set).
I will show how this very general principle fails when agents being classified strategically respond to the classifier, and present an intuitive fix that leads to provable (and in fact, optimal) generalization guarantees under strategic manipulation. I will then discuss the role of incentive-compatibility in strategic classification, and present experimental results that illustrate how the theoretical results can guide practice.
Vincent Conitzer (Chair)
Nika Haghtalab (University of California, Berkeley)
Vahab Mirrokni (Google Research)
Renato Paes Leme (Google Research)