Mechanism Design for Cost Optimal PAC Learning in the Presence of Strategic Noisy Annotators
This work addresses the challenge of learning from unreliable human annotators in machine learning, with incremental contributions in mechanism design for strategic settings.
The paper tackles the problem of PAC learning from multiple noisy annotators by deriving sample complexity bounds for a known algorithm under complete information and designing a cost-optimal procurement auction mechanism to elicit true noise rates from strategic annotators under incomplete information.
We consider the problem of Probably Approximate Correct (PAC) learning of a binary classifier from noisy labeled examples acquired from multiple annotators (each characterized by a respective classification noise rate). First, we consider the complete information scenario, where the learner knows the noise rates of all the annotators. For this scenario, we derive sample complexity bound for the Minimum Disagreement Algorithm (MDA) on the number of labeled examples to be obtained from each annotator. Next, we consider the incomplete information scenario, where each annotator is strategic and holds the respective noise rate as a private information. For this scenario, we design a cost optimal procurement auction mechanism along the lines of Myerson's optimal auction design framework in a non-trivial manner. This mechanism satisfies incentive compatibility property, thereby facilitating the learner to elicit true noise rates of all the annotators.