New approach optimizes policies in adversarial MDPs using adversarial learning.
problem Optimizing policies in adversarial Markov decision processes.
method Adversarial learning on advantage functions, extending previous reductions.
result Stronger regret criteria and performance guarantees for policy optimization.
Study robust regression learning under adversarial attacks.
problem Understanding which function classes are learnable in the presence of adversarial attacks.
method Introduced a novel agnostic sample compression scheme and used fat-shattering dimension to construct adversarially robust sample compression schemes.
result Finite fat-shattering dimension classes are learnable in both realizable and agnostic settings.
Geometric approach improves reinforcement learning representation.
problem Improving reinforcement learning representation learning.
method Formal evidence through geometric properties of value functions.
result Optimizing value functions reduces to predicting adversarial value functions (AVFs).
Adversarial examples have been shown to exist for a variety of deep learning architectures. Deep reinforcement learning has shown promising results on training agent policies directly on raw inputs such as image pixels. In this paper we present a novel study into adversarial attacks on deep reinforcement learning polic…
Unified approach for robust low rank matrix estimation with adversaries.
problem Robust low rank matrix estimation in the presence of adversaries.
method Unified approach combining Huber loss and nuclear norm penalization.
result Sharp estimation error bounds for matrix compressed sensing and completion.
This paper analyzes how machine learning models resist adversarial attacks in nonparametric regression.
problem Adversarial attacks on machine learning models in nonparametric regression.
method Theoretical analysis of minimax rates of convergence under adversarial sup-norm.
result The minimax rate under adversarial attacks is the sum of two terms: standard rate and deviation of true function.
A new method generates natural-looking adversarial examples by bounding internal activation values.
problem Creating natural-looking adversarial examples that closely mimic the original input.
method Bounding internal activation values through a distribution quantile bound and polynomial barrier loss function.
result Our attack achieves similar success and confidence levels as state-of-the-art methods but with more natural-looking perturbations.
A new imputation method MissARF uses adversarial random forests for fast and accurate missing value imputation.
problem Handling missing values in biostatistical analyses.
method Adversarial Random Forests (ARF) for density estimation and data synthesis.
result MissARF performs comparably to state-of-the-art methods in imputation quality and runtime.
Paper optimizes learning MDPs with adversarial rewards using linear models.
problem Learning MDPs with adversarial rewards and changing reward functions.
method Proposes POWERS, an optimistic policy optimization algorithm.
result Achieves nearly optimal regret bound of O ( d H T ) O(dH\sqrt{T}) O ( d H T ) . This paper improves continuous adversarial training for LLMs using in-context learning theory.
problem Efficiently defending large language models (LLMs) against jailbreak attacks.
method The paper presents a theoretical analysis of continuous adversarial training (CAT) for LLMs based on in-context learning (ICL) theory, proving a robust generalization bound and proposing an improved regularization term.
result The robust generalization bound explains why CAT can defend against jailbreak prompts and shows that LLM robustness is related to embedding matrix singular values.
Oracle-efficient algorithms for online learning with smoothed and hint-adversaries.
problem Online learning with beyond worst-case adversaries.
method Oracle-efficient algorithms for two settings: smoothed analysis and K K K -hint transductive learning. result Oracle-efficient regret bounds for learning real-valued and binary-valued functions.
Detects adversarial examples using SHAP values from neural networks.
problem Vulnerability of deep neural networks to adversarial attacks.
method Uses SHAP values from internal layers to distinguish normal from adversarial inputs.
result Demonstrates high detection accuracy and strong generalization to various attack methods.
Malware is constantly adapting in order to avoid detection. Model based malware detectors, such as SVM and neural networks, are vulnerable to so-called adversarial examples which are modest changes to detectable malware that allows the resulting malware to evade detection. Continuous-valued methods that are robust to a…
FE-GAN improves VaR and ES estimation in financial risk management.
problem Improving VaR and ES estimation in financial risk management.
method Feature-Enriched Generative Adversarial Networks (FE-GAN) with specialized models like WGAN and Tail-GAN.
result FE-GAN significantly outperforms traditional GANs in VaR and ES estimation.
New sparsity attacks degrade DNN efficiency, raising concerns for resource-constrained systems.
problem Vulnerabilities in DNNs through energy and latency attacks.
method Proposed sparsity attacks that modify DNN inputs to reduce activation sparsity, increasing execution time and energy consumption.
result Adversarial sparsity attacks can degrade DNN efficiency by up to 1.82x in image recognition DNNs.
RORL improves offline RL robustness with conservative smoothing.
problem Distribution shift and robustness issues in offline RL.
method RORL introduces regularization and conservative smoothing for robustness.
result RORL achieves state-of-the-art performance and robustness to adversarial perturbations.
Paper introduces SPADE method to protect classifiers from OOD and adversarial samples.
problem Protecting classifiers from out-of-distribution and adversarial samples.
method SPADE method based on GEV model in latent space.
result Provable protection against OOD and adversarial samples.
Discretizing input space improves DLN robustness against adversarial attacks.
problem Improving machine learning models' resistance to adversarial attacks.
method Input discretization and Binary Neural Networks (BNNs).
result 2-bit input discretization significantly enhances adversarial robustness with minimal accuracy loss.
New method improves neural network robustness to adversarial attacks.
problem Improving adversarial robustness of neural networks.
method Inspired by adaptive control theory, the approach uses persistency of excitation to constrain gradient descent updates.
result Networks trained with the PoE-motivated learning rate schedule are significantly more robust to adversarial attacks.
New framework for efficient query-based imitation learning.
problem Aligning agent policy with human expert behavior without prior knowledge.
method Adversarial reward query with successor representation.
result Significantly outperforms uncertainty-based methods in query efficiency.
AGAC uses an adversary to enhance exploration in complex tasks.
problem Sample inefficiency in complex environments, especially in tasks requiring efficient exploration.
method Integrates an adversary into the actor-critic framework to encourage innovative exploration strategies.
result AGAC leads to more exhaustive exploration and outperforms state-of-the-art methods.
Guarantees adversarial robustness of Gaussian Process Classification models.
problem Ensuring robustness of Gaussian Process Classification models against adversarial attacks.
method Developed a branch and bound optimisation algorithm to compute minimum and maximum classification probabilities over a compact subset of input space.
result Guaranteed to find values ε-close to actual robustness values in finitely many iterations for any error threshold ε.
New algorithms reduce regret in adversarial linear contextual bandits.
problem Adversarial linear contextual bandits with changing loss functions.
method Developed two algorithms: RealLinExp3 and RobustLinExp3.
result Achieved optimal regret bounds for the first time.
We consider a model of robust learning in an adversarial environment. The learner gets uncorrupted training data with access to possible corruptions that may be affected by the adversary during testing. The learner's goal is to build a robust classifier, which will be tested on future adversarial examples. The adversar…
AR-A3C improves A3C's robustness to noisy environments.
problem Neural networks are vulnerable to noise from unexpected sources in reinforcement learning.
method Introduced an adversarial agent to make the learning process more robust.
result AR-A3C outperforms A3C in both clean and noisy environments.
New neural network method hides input information in complex-valued features to protect privacy.
problem Preventing adversaries from inferring input attributes from neural network features.
method Transforming real-valued features into complex-valued ones, making input hidden in a randomized phase.
result Significantly diminishes adversary's ability to infer input while preserving high accuracy.
Study adversarial classification with data corruption up to ε, deriving geometric flows.
problem Optimizing classifiers against adversarial data corruption.
method Variational analysis, geometric flows, mean curvature equations.
result Rigorous proof of initial value problem for small ε, global minimizer of adversarial problem.
We formulate a private learning model to study an intrinsic tradeoff between privacy and query complexity in sequential learning. Our model involves a learner who aims to determine a scalar value, v ∗ v^* v ∗ , by sequentially querying an external database and receiving binary responses. In the meantime, an adversary observes…
A technique finds adversarial examples for deep neural networks using human perception.
problem Finding adversarial examples for deep neural networks without access to internal structure.
method Covariance Matrix Adaptation Evolution Strategy (CMA-ES) with perception-in-the-loop.
result CMA-ES can find adversarial examples with human feedback, showing favorable performance.
New algorithm identifies near-optimal policies in adversarial distributed RL settings.
problem Adversarial agents in distributed RL settings that can collude and report arbitrary data.
method Weighted-Clique algorithm for robust mean estimation from batches, combined with novel distributed algorithms.
result Achieves superior robustness guarantees and near-optimal sample complexities in both offline and online settings.
Study robust learning of Lipschitz functions under corrupted binary signals.
problem Learning a Lipschitz function with corrupted binary signals in a context of unknown corruption rounds.
method Introduced agnostic checking and new analysis techniques to design algorithms for symmetric and pricing losses.
result Achieved small cumulative loss for both symmetric and pricing losses.
New framework for fairness in machine learning models using SHAP values and adversarial learning.
problem Fairness of model predictions, especially for unprivileged groups.
method Develops a new fairness definition and a framework using SHAP values and adversarial learning to mitigate bias.
result Models produced are fairer and performant, demonstrating the approach on various datasets.
Paper bounds convergence rate of adversarial surrogate risk.
problem Vulnerability of binary classification models to adversarial attacks.
method Characterizes conditions for adversarial consistency and provides surrogate risk bounds.
result Surrogate risk bounds quantify the rate of convergence of adversarial classification risk.
New geometric interpretation explains over-parameterized models and adversarial perturbations.
problem Geometric understanding of over-parameterized regression and adversarial perturbations.
method Alternative geometric interpretation of regression in feature space.
result Adversarial perturbations are a natural feature of biased models due to underlying geometry.
New estimator robust to adversarial noise and data heterogeneity.
problem Sensitive to adversarial noise and poor performance with heterogeneous data.
method Distributionally robust estimator minimizing worst-case conditional expected loss over adversarial distributions.
result Efficiently finds non-parametric local estimates via convex optimization.
Study efficient reinforcement learning for partially observed systems with linear structure.
problem Efficient reinforcement learning for partially observed Markov decision processes with linear structure.
method Proposes OP-TENET algorithm using a Bellman operator with finite memory, adversarial integral equation, and optimistic exploration.
result Achieves ε-optimal policy within O(1/ε^2) episodes with polynomial sample complexity in intrinsic dimension.
Study on robustness of subspace learning from adversarial modifications.
problem Adversarial robustness of subspace learning problems.
method Characterization of optimal rank-one attack strategies and generalization to non-rank-constrained scenarios.
result Optimal attack strategies depend on singular values and adversary's energy budget.
Motivated by online advertising auctions, we consider repeated Vickrey auctions where goods of unknown value are sold sequentially and bidders only learn (potentially noisy) information about a good's value once it is purchased. We adopt an online learning approach with bandit feedback to model this problem and derive …
Improves adversarial robustness by constraining logits with a bounded function.
problem Improving adversarial robustness in deep learning models.
method Addition of a bounded function before softmax to constrain logits.
result Our method improves adversarial robustness without requiring adversarial training.
IFGAN uses feature-specific GANs for missing value imputation.
problem Missing value imputation in data mining.
method Feature-specific Generative Adversarial Networks (GAN).
result IFGAN outperforms state-of-the-art algorithms in various missing conditions.
Study improves adversarial classification using distributionally robust models.
problem Improving robustness against adversarial attacks in classification models.
method Distributionally robust chance constraints with Wasserstein ambiguity, reformulated as a regularized ramp loss minimization problem.
result Standard descent methods can converge to the global minimizer for the distributionally robust adversarial classification model.
We augment adversarial training (AT) with worst case adversarial training (WCAT) which improves adversarial robustness by 11% over the current state-of-the-art result in the ℓ 2 \ell_2 ℓ 2 norm on CIFAR-10. We obtain verifiable average case and worst case robustness guarantees, based on the expected and maximum values of the…
A novel approach uses an ensemble of Gaussian processes for robust and adaptive reinforcement learning.
problem Adaptive reinforcement learning in large or continuous state spaces.
method Online scalable (OS) approach with a weighted ensemble of Gaussian processes.
result The ensemble approach improves performance in adversarial settings.
The paper addresses learner privacy in convex optimization with feedback.
problem Privacy risks from eavesdropping adversaries observing learner's queries.
method Optimally obfuscating learner's queries to make their learned optimal value hard to estimate.
result Query complexity overhead is additive in L L L in the minimax formulation, multiplicative in L L L in the Bayesian formulation. Deep, wide ConvResNets can approximate functions and their smoothness.
problem Function approximation and smoothness in deep networks.
method Analyzing ConvResNets, proving their ability to approximate functions and their smoothness.
result Large ConvResNets can approximate functions and exhibit sufficient first-order smoothness.
Paper tackles constrained robust optimization without gradients.
problem Constrained robust (min-max) optimization in a black-box setting.
method Integrates ZO gradient estimator with alternating projected SGD-ascent method.
result Proposed ZO-Min-Max framework converges sub-linearly under mild conditions.
Bayesian Robust Optimization for Imitation Learning (BROIL) balances risk and reward.
problem Learning robust policies for new states in imitation learning.
method Bayesian reward function inference and user-specific risk tolerance.
result BROIL outperforms risk-sensitive and risk-neutral algorithms.
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…