Differentially Private Policy Evaluation

2016/03/07 by Borja Balle, Balle, Borja, Maziar Gomrokchi +3 · 3 citations
Computer Science · Mathematics · #Age of Information Optimization #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Mobile Crowdsensing and Crowdsourcing #Privacy-Preserving Technologies in Data #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1603.02010

arxiv created 2016/03/07 · openalex publication_date 2016/03/07 · arxiv updated 2016/03/08 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We present the first differentially private algorithms for reinforcement learning, which apply to the task of evaluating a fixed policy. We establish two approaches for achieving differential privacy, provide a theoretical analysis of the privacy and utility of the two algorithms, and show promising results on simple empirical examples.

Cited by

• Multi-level Certified Defense Against Poisoning Attacks in Offline Reinforcement Learning
• Improved Algorithms for Differentially Private Language Model Alignment
• Deep Reinforcement Learning: An Overview

Related

• Privacy Preserving Off-Policy Evaluation
• Differentially Private Policy Gradient
• Near-Optimal Differentially Private Reinforcement Learning
• Differentially Private Deep Model-Based Reinforcement Learning
• Private Reinforcement Learning with PAC and Regret Guarantees