Privacy-Preserving Bandits

TL;DR

This paper introduces Privacy-Preserving Bandits (P2B), a system that enables local agents to learn from each other in a privacy-preserving way, maintaining competitive recommendation accuracy and click-through rates.

Contribution

The paper proposes P2B, a novel system that combines differential privacy with multi-agent collaboration for personalized bandit algorithms on user devices.

Findings

Only 2.6% and 3.6% accuracy decrease in synthetic benchmarks.

A 0.0025 CTR increase in real-world online advertising.

Effective privacy-preserving personalization with minimal performance loss.

Abstract

Contextual bandit algorithms~(CBAs) often rely on personal data to provide recommendations. Centralized CBA agents utilize potentially sensitive data from recent interactions to provide personalization to end-users. Keeping the sensitive data locally, by running a local agent on the user's device, protects the user's privacy, however, the agent requires longer to produce useful recommendations, as it does not leverage feedback from other users. This paper proposes a technique we call Privacy-Preserving Bandits (P2B); a system that updates local agents by collecting feedback from other local agents in a differentially-private manner. Comparisons of our proposed approach with a non-private, as well as a fully-private (local) system, show competitive performance on both synthetic benchmarks and real-world data. Specifically, we observed only a decrease of 2.6% and 3.6% in multi-label classification accuracy, and a CTR increase of 0.0025 in online advertising for a privacy budget $\epsilon \approx 0.693$. These results suggest P2B is an effective approach to challenges arising in on-device privacy-preserving personalization.