arXiv research

New algorithms for planning with adversarial changes in costs.

This paper proposes CSADA to make DNNs cost-sensitive.

Several recent works have developed methods for training classifiers that are certifiably robust against norm-bounded adversarial perturbations. These methods assume that all the adversarial transformations are equally important, which is seldom the case in real-world applications. We advocate for cost-sensitive robust…

Machine-learning models for security-critical applications such as bot, malware, or spam detection, operate in constrained discrete domains. These applications would benefit from having provable guarantees against adversarial examples. The existing literature on provable adversarial robustness of models, however, exclu…

This paper tackles near-optimal adversarial RL with switching costs, providing algorithms and matching lower bounds.

We study the power of different types of adaptive (nonoblivious) adversaries in the setting of prediction with expert advice, under both full-information and bandit feedback. We measure the player's performance using a new notion of regret, also known as policy regret, which better captures the adversary's adaptiveness…

Algorithm for bandits with switching costs achieves optimal regret bounds.

Adversarial training, in which a network is trained on adversarial examples, is one of the few defenses against adversarial attacks that withstands strong attacks. Unfortunately, the high cost of generating strong adversarial examples makes standard adversarial training impractical on large-scale problems like ImageNet…

New interpretation of OT regularization as adversarial ground cost.

Motivated by safety-critical classification problems, we investigate adversarial attacks against cost-sensitive classifiers. We use current state-of-the-art adversarially-resistant neural network classifiers [1] as the underlying models. Cost-sensitive predictions are then achieved via a final processing step in the fe…

Study cyber-attacks on RL algorithms, focusing on cost signal manipulation.

This paper investigates the adversarial Bandits with Knapsack (BwK) online learning problem, where a player repeatedly chooses to perform an action, pays the corresponding cost, and receives a reward associated with the action. The player is constrained by the maximum budget $B$ that can be spent to perform actions, an…

Study examines trading costs on Uniswap, finding adversarial slippage is significant for large trades and certain assets.

Adversarial training, in which a network is trained on both adversarial and clean examples, is one of the most trusted defense methods against adversarial attacks. However, there are three major practical difficulties in implementing and deploying this method - expensive in terms of extra memory and computation costs; …

Study statistical guarantees for DRO with OT and OT-regularized divergences.

New method minimizes regret in AMDP with high probability.

New algorithm controls linear systems with bandit feedback, achieving optimal regret.

New algorithm achieves logarithmic regret for adversarial online control.

New method certifies joint adversarial robustness of model ensembles.

I describe an optimal control view of adversarial machine learning, where the dynamical system is the machine learner, the input are adversarial actions, and the control costs are defined by the adversary's goals to do harm and be hard to detect. This view encompasses many types of adversarial machine learning, includi…

New framework improves adversarial robustness in one-stage L2D.

New method enhances adversarial robustness of deep learning models.

In recent years, deep neural networks have demonstrated outstanding performance in many machine learning tasks. However, researchers have discovered that these state-of-the-art models are vulnerable to adversarial examples: legitimate examples added by small perturbations which are unnoticeable to human eyes. Adversari…

A new algorithm finds minimizers in dueling optimization with a monotone adversary.

In this work we consider adversarial contextual bandits with risk constraints. At each round, nature prepares a context, a cost for each arm, and additionally a risk for each arm. The learner leverages the context to pull an arm and then receives the corresponding cost and risk associated with the pulled arm. In additi…

DL models for MTS regression are vulnerable to adversarial attacks, posing risks in safety-critical applications.

In many real-world applications of Machine Learning it is of paramount importance not only to provide accurate predictions, but also to ensure certain levels of robustness. Adversarial Training is a training procedure aiming at providing models that are robust to worst-case perturbations around predefined points. Unfor…

New method reduces regret for sparse adversarial SSP problems.

Game theory helps machine learn better from adversarial queries.

While progress has been made in understanding the robustness of machine learning classifiers to test-time adversaries (evasion attacks), fundamental questions remain unresolved. In this paper, we use optimal transport to characterize the minimum possible loss in an adversarial classification scenario. In this setting, …

TRASHFIRE improves model robustness by analyzing training rates and costs.

BPN defends against adversarial attacks by generating beneficial perturbations.

Study optimal transport for robust optimization, showing how adversary's strategy relates to regularization.

AdvMind detects adversary intent in black-box attacks with high accuracy.

New algorithms tackle adversarial combinatorial bandits with switching costs.

Optimizes query routing to LLMs under cost and resource constraints.

Paper analyzes adversarial risk using optimal transport.

Paper proposes an adversarial sampling method for efficient extreme classification.

Generalised regularisation equals robustness for exotic function classes.

We investigate the use of entropy-regularized optimal transport (EOT) cost in developing generative models to learn implicit distributions. Two generative models are proposed. One uses EOT cost directly in an one-shot optimization problem and the other uses EOT cost iteratively in an adversarial game. The proposed gene…

New method shows adversarial training can be as effective as traditional training without the high cost.

The paper identifies potential adversarial samples near decision boundaries of neural networks.

Study adversarial attacks on automated trading systems.

Efficient algorithm controls unknown systems with adversarial perturbations.

Adversarial training improves audio classifier robustness.

The security of Deep Reinforcement Learning (Deep RL) algorithms deployed in real life applications are of a primary concern. In particular, the robustness of RL agents in cyber-physical systems against adversarial attacks are especially vital since the cost of a malevolent intrusions can be extremely high. Studies hav…

Adaptive Divergence speeds up AO for heavy generators.

New attack manipulates UCB algorithm, new defense algorithm reduces pseudo-regret.