2 papers
cs.GT2024
Strategyproof Learning with Advice
Eric Balkanski, Cherlin Zhu
An important challenge in robust machine learning is when training data is provided by strategic sources who may intentionally report erroneous data for their own benefit. A line o…
cs.GT2023
Online Mechanism Design with Predictions
Eric Balkanski, Vasilis Gkatzelis, Xizhi Tan +1
Aiming to overcome some of the limitations of worst-case analysis, the recently proposed framework of "algorithms with predictions" allows algorithms to be augmented with a (possib…