Adversarial Machine Learning: Bayesian Perspectives

Abstract

Adversarial Machine Learning (AML) is emerging as a major field aimed at protecting machine learning (ML) systems against security threats: in certain scenarios there may be adversaries that actively manipulate input data to fool learning systems. This creates a new class of security vulnerabilities that ML systems may face, and a new desirable property called adversarial robustness essential to trust operations based on ML outputs. Most work in AML is built upon a game-theoretic modelling of the conflict between a learning system and an adversary, ready to manipulate input data. This assumes that each agent knows their opponent's interests and uncertainty judgments, facilitating inferences based on Nash equilibria. However, such common knowledge assumption is not realistic in the security scenarios typical of AML. After reviewing such game-theoretic approaches, we discuss the benefits that Bayesian perspectives provide when defending ML-based systems. We demonstrate how the Bayesian approach allows us to explicitly model our uncertainty about the opponent's beliefs and interests, relaxing unrealistic assumptions, and providing more robust inferences. We illustrate this approach in supervised learning settings, and identify relevant future research problems.

References in corpus (21)

• Explaining and Harnessing Adversarial Examples
• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Towards Deep Learning Models Resistant to Adversarial Attacks
• Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning
• Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples
• On Evaluating Adversarial Robustness
• Security Evaluation of Pattern Classifiers under Attack
• Technical Report on the CleverHans v2.1.0 Adversarial Examples Library
• On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Models
• Learning Representations for Counterfactual Inference
• Adv-BNN: Improved Adversarial Defense through Robust Bayesian Neural Network
• Tactics of Adversarial Attack on Deep Reinforcement Learning Agents
• A Survey of Game Theoretic Approaches for Adversarial Machine Learning in Cybersecurity Tasks
• Adversarial classification: An adversarial risk analysis approach
• A Causal View on Robustness of Neural Networks
• Stochastic Gradient MCMC with Repulsive Forces
• Model-based Reinforcement Learning for Semi-Markov Decision Processes with Neural ODEs
• Augmented Probability Simulation Methods for Sequential Games
• Query-Efficient and Scalable Black-Box Adversarial Attacks on Discrete Sequential Data via Bayesian Optimization
• Adversarial attacks against Bayesian forecasting dynamic models
• Knowing When to Stop: Evaluation and Verification of Conformity to Output-size Specifications

Cited by in corpus (1)

• Manipulating hidden-Markov-model inferences by corrupting batch data