Privacy-Preserving Bandits
This addresses privacy concerns in on-device personalization for users, though it is incremental as it builds on existing bandit and differential privacy methods.
The paper tackles the problem of balancing privacy and performance in contextual bandit algorithms for recommendations by proposing Privacy-Preserving Bandits (P2B), which updates local agents with differentially-private feedback from other users, resulting in only a 2.6-3.6% decrease in accuracy and a 0.0025 increase in CTR for a privacy budget of ε≈0.693.
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 $ε\approx 0.693$. These results suggest P2B is an effective approach to challenges arising in on-device privacy-preserving personalization.