arXiv research

Mathematical conditions and practical computations for adversarial robustness measures are established.

New method for efficient proximal mapping of 1-path-norm in shallow networks.

Algorithm minimizes regret and converges to equilibria in Markov games.

As adversarial attacks pose a serious threat to the security of AI system in practice, such attacks have been extensively studied in the context of computer vision applications. However, few attentions have been paid to the adversarial research on automatic path finding. In this paper, we show dominant adversarial exam…

Recently, researchers have started decomposing deep neural network models according to their semantics or functions. Recent work has shown the effectiveness of decomposed functional blocks for defending adversarial attacks, which add small input perturbation to the input image to fool the DNN models. This work proposes…

Study online learning in MDPs with aggregate bandit feedback, achieving low regret in both stochastic and adversarial settings.

New algorithms for planning with adversarial changes in costs.

New method reduces regret for sparse adversarial SSP problems.

SAM improves deep learning by relaxing Bayes objective.

Generative Adversarial Networks have been shown to be powerful in generating content. To this end, they have been studied intensively in the last few years. Nonetheless, training these networks requires solving a saddle point problem that is difficult to solve and slowly converging. Motivated from techniques in the reg…

New betting strategy reduces regret to ln(ln n) with protection against adversarial data.

This paper describes and evaluates the use of Generative Adversarial Networks (GANs) for path planning in support of smart mobility applications such as indoor and outdoor navigation applications, individualized wayfinding for people with disabilities (e.g., vision impairments, physical disabilities, etc.), path planni…

Given two points on a soup can or conical cup with lid, we find and classify all paths of minimal length connecting them. When the number of minimal paths is finite, there are at most four on a can and three on a cup. At worst, minimal paths are piece-wise smooth with three components, each of which is a classical geod…

Proposes using mode connectivity to improve adversarial robustness of neural networks.

2DSig-Detect detects adversarial perturbations in images.

LSDAT reduces query efficiency for decision-based adversarial attacks.

Adversarial consistency depends on the uniqueness of adversarial Bayes classifiers.

We study adaptive regret bounds in terms of the variation of the losses (the so-called path-length bounds) for both multi-armed bandit and more generally linear bandit. We first show that the seemingly suboptimal path-length bound of (Wei and Luo, 2018) is in fact not improvable for adaptive adversary. Despite this neg…

In this paper we consider the length minimizing properties of Hamiltonian paths generated by quasi-autonomous Hamiltonians on symplectically aspherical manifolds. Motivated by the work of L. Polterovich and M. Schwarz, we study the role of the fixed global extrema in the Floer complex of the generating Hamiltonian. Our…

Characterizes paths minimizing anisotropic lengths in Euclidean space.

Algorithm minimizes risk for multiclass classification of stochastic diffusion paths.

Paper bounds convergence rate of adversarial surrogate risk.

We use the criteria of Lalonde and McDuff to determine a new class of examples of length minimizing paths in the group $Ham(M)$. For a compact symplectic manifold $M$ of dimension two or four, we show that a path in $Ham(M)$, generated by an autonomous Hamiltonian and starting at the identity, which induces no non-cons…

In this paper we first show that the necessary condition introduced in our previous paper is also a sufficient condition for a path to be a geodesic in the group $\Ham^c(M)$ of compactly supported Hamiltonian symplectomorphisms. This applies with no restriction on $M$. We then discuss conditions which guarantee that su…

Tricks adversarial attacks to target specific classes, improving classifier accuracy.

Diagonal linear networks converge to lasso regularization path during training.

The paper proposes a neural network method to calibrate LSV models without interpolation.

Adversarial example generation becomes a viable method for evaluating the robustness of a machine learning model. In this paper, we consider hard-label black-box attacks (a.k.a. decision-based attacks), which is a challenging setting that generates adversarial examples based on only a series of black-box hard-label que…

Paper introduces non-adversarial training for Neural SDEs using signature kernel scores.

We solve the paradox of score-based methods by minimizing path variance.

This work proposes SDI regularization to improve adversarial robustness.

Proposes neuron alignment to optimize mode connectivity in neural networks.

We consider small-time asymptotics for diffusion processes conditioned by their initial and final positions, under the assumption that the diffusivity has a sub-Riemannian structure, not necessarily of constant rank. We show that, if the endpoints are joined by a unique path of minimal energy, and lie outside the sub-R…

New geometric insights reveal properties of adversarial training problems.

Study improves adversarial classification using distributionally robust models.

New analysis of annealing paths in sampling and estimation.

Neural SDEs improve time series generation efficiency.

This study connects Jacobian regularization to adversarial robustness and improves generalization.

Neural Jump ODEs model Itô processes without adversarial training.

The recently developed bag-of-paths (BoP) framework consists in setting a Gibbs-Boltzmann distribution on all feasible paths of a graph. This probability distribution favors short paths over long ones, with a free parameter (the temperature $T$) controlling the entropic level of the distribution. This formalism enables…

This paper optimizes paths for generative models using kinetic energy.

New algorithms reduce regret in online MDPs by adapting to data and variance.

The paper shows how policy regularization acts like an adversary to improve robustness.

Sub-Riemannian geometry connects bike paths to mathematical curves.

The paper proves a regret bound for a sub-Gaussian mixture on unbounded data.

Most previous works usually explained adversarial examples from several specific perspectives, lacking relatively integral comprehension about this problem. In this paper, we present a systematic study on adversarial examples from three aspects: the amount of training data, task-dependent and model-specific factors. Pa…

Endogenous randomness emerges from adversarial market learning.

It has been suggested that adversarial examples cause deep learning models to make incorrect predictions with high confidence. In this work, we take the opposite stance: an overly confident model is more likely to be vulnerable to adversarial examples. This work is one of the most proactive approaches taken to date, as…