Strategyproof Learning with Advice

TL;DR

This paper explores strategyproof machine learning algorithms that balance accuracy when advice is correct with robustness against erroneous data, providing optimal tradeoffs for certain function classes.

Contribution

It introduces the first non-trivial tradeoffs between consistency and robustness in strategyproof regression and classification, with optimal bounds for simple function classes.

Findings

Deterministic, strategyproof mechanism for constant functions with optimal tradeoff

Extension of mechanisms to homogeneous linear regression

Impossibility results for multi-label classification with three or more labels

Abstract

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 of work at the intersection of machine learning and mechanism design aims to deter strategic agents from reporting erroneous training data by designing learning algorithms that are strategyproof. Strategyproofness is a strong and desirable property, but it comes at a cost in the approximation ratio of even simple risk minimization problems. In this paper, we study strategyproof regression and classification problems in a model with advice. This model is part of a recent line on mechanism design with advice where the goal is to achieve both an improved approximation ratio when the advice is correct (consistency) and a bounded approximation ratio when the advice is incorrect (robustness). We provide the first non-trivial consistency-robustness tradeoffs for strategyproof regression and classification, which hold for simple yet interesting classes of functions. For classes of constant functions, we give a deterministic and strategyproof mechanism that is, for any $\gamma \in (0, 2]$, $1+\gamma$ consistent and $1 + 4/\gamma$ robust and provide a lower bound that shows that this tradeoff is optimal. We extend this mechanism and its guarantees to homogeneous linear regression over $\mathbb{R}$. In the binary classification problem of selecting from three or more labelings, we present strong impossibility results for both deterministic and randomized mechanism. Finally, we provide deterministic and randomized mechanisms for selecting from two labelings.