Strategyproof Classification under Constant Hypotheses: A Tale of Two Functions

This work designs mechanisms (both deterministic and randomized) that reach an approximately optimal decision and are strategyproof, i.e., agents are best off when they tell the truth.