arXiv research

Study on hypothesis testing games with adversarial classification, showing convergence rates.

This paper examines various definitions of adversarial risk and their implications.

Game theory helps machine learn better from adversarial queries.

New framework crafts adversarial examples for neural networks.

Paper explains adversarial training's robust overfitting through a minimax game perspective.

Randomized classifiers outperform deterministic ones in robustness against adversarial attacks.

Study on learning strategies in adaptive Markov games with policy regret as metric.

We study the problem of learning classifiers robust to universal adversarial perturbations. While prior work approaches this problem via robust optimization, adversarial training, or input transformation, we instead phrase it as a two-player zero-sum game. In this new formulation, both players simultaneously play the s…

Survey of game theory methods to secure machine learning against adversarial attacks.

Gradient methods solve Stackelberg games in high-dimensional, continuous decision spaces.

Paper analyzes adversarial attacks and defenses using game theory.

Neural networks are known to be vulnerable to adversarial examples. Carefully chosen perturbations to real images, while imperceptible to humans, induce misclassification and threaten the reliability of deep learning systems in the wild. To guard against adversarial examples, we take inspiration from game theory and ca…

Games generalize the single-objective optimization paradigm by introducing different objective functions for different players. Differentiable games often proceed by simultaneous or alternating gradient updates. In machine learning, games are gaining new importance through formulations like generative adversarial netwo…

New theorem guarantees approximate equilibrium in non-convex games.

Improves GANs training through game theory.

Study adversarial attacks on cost-sensitive classifiers.

We present a new autoencoder-type architecture that is trainable in an unsupervised mode, sustains both generation and inference, and has the quality of conditional and unconditional samples boosted by adversarial learning. Unlike previous hybrids of autoencoders and adversarial networks, the adversarial game in our ap…

Optimal adversarial noise algorithms for multiple classifiers using game theory.

Proposes a new training algorithm for zero-sum games to avoid convergence issues.

Study optimal policy regret in partially observable Markov games with adaptive opponents.

New method UADs improves transferability of adversarial perturbations.

This work tackles learning Markov games with adversarial opponents and achieves both average reward and exploitation.

We describe an approximate dynamic programming (ADP) approach to compute approximations of the optimal strategies and of the minimal losses that can be guaranteed in discounted repeated games with vector-valued losses. Such games prominently arise in the analysis of regret in repeated decision-making in adversarial env…

Gradient-descent-ascent dynamics can exhibit various behaviors in non-convex non-concave games.

New measure of policy regret shows compatibility with traditional external regret in adversarial games.

New tools understand and control dynamics in n-player differentiable games.

Despite the improved accuracy of deep neural networks, the discovery of adversarial examples has raised serious safety concerns. In this paper, we study two variants of pointwise robustness, the maximum safe radius problem, which for a given input sample computes the minimum distance to an adversarial example, and the …

Paper improves autonomous vehicle safety and efficiency with new reinforcement learning methods.

Bayesian methods improve adversarial machine learning robustness.

Gradient-based methods for games suffer from discrete update steps that cause drift, affecting performance.

Gradient Descent Ascent converges to von-Neumann solution in hidden zero-sum games.

New method improves convergence for smooth games.

With a large number of sensors and control units in networked systems, distributed support vector machines (DSVMs) play a fundamental role in scalable and efficient multi-sensor classification and prediction tasks. However, DSVMs are vulnerable to adversaries who can modify and generate data to deceive the system to mi…

DefogGAN predicts hidden RTS game information to aid strategic decision-making.

Generative Adversarial Networks (GAN) have become one of the most successful frameworks for unsupervised generative modeling. As GANs are difficult to train much research has focused on this. However, very little of this research has directly exploited game-theoretic techniques. We introduce Generative Adversarial Netw…

This paper examines how adversarial perturbations affect model performance and equilibrium learning.

Partial monitoring is a generalization of the well-known multi-armed bandit framework where the loss is not directly observed by the learner. We complete the classification of finite adversarial partial monitoring to include all games, solving an open problem posed by Bartok et al. [2014]. Along the way we simplify and…

New algorithms improve on bandit feedback in matrix games with unknown payoff matrices.

This work simplifies adversarial attacks using neural networks, reducing computation and improving training convergence.

Despite the considerable success enjoyed by machine learning techniques in practice, numerous studies demonstrated that many approaches are vulnerable to attacks. An important class of such attacks involves adversaries changing features at test time to cause incorrect predictions. Previous investigations of this proble…

Introduces SM-games to analyze machine learning interactions.

Algorithm reduces regret in complex game scenarios.

Generative models use EOT cost for better image generation.

New metrics fail adversarial tests, with some more robust than others.

The paper develops algorithms for competitive RL in partially observable MGs.

New method learns optimal prediction strategies in adversarial games.

ARL makes market makers resilient to adversarial conditions.

We present a data-driven framework called generative adversarial privacy (GAP). Inspired by recent advancements in generative adversarial networks (GANs), GAP allows the data holder to learn the privatization mechanism directly from the data. Under GAP, finding the optimal privacy mechanism is formulated as a constrain…