A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
New method stabilizes private LASSO for high-dimensional data with diverse covariate scales.
Machine learning models have been shown vulnerable to adversarial attacks launched by adversarial examples which are carefully crafted by attacker to defeat classifiers. Deep learning models cannot escape the attack either. Most of adversarial attack methods are focused on success rate or perturbations size, while we a…
Given the ability to directly manipulate image pixels in the digital input space, an adversary can easily generate imperceptible perturbations to fool a Deep Neural Network (DNN) image classifier, as demonstrated in prior work. In this work, we propose ShapeShifter, an attack that tackles the more challenging problem o…
Node-perturbation learning is a type of statistical gradient descent algorithm that can be applied to problems where the objective function is not explicitly formulated, including reinforcement learning. It estimates the gradient of an objective function by using the change in the object function in response to the per…
Study privacy and accuracy in high-dimensional LASSO with perturbation mechanisms.
Within the unmanageably large class of nonconvex optimization, we consider the rich subclass of nonsmooth problems that have composite objectives---this already includes the extensively studied convex, composite objective problems as a special case. For this subclass, we introduce a powerful, new framework that permits…
This paper analyzes privacy-preserving methods for sparse model optimization.
Enhances safety of 3D object detection neural networks.
In this paper, we address the issue of linear stability of Schwarzschild space- time subject to certain axisymmetric perturbations. In particular, we prove that associ- ated solutions to the linearized vacuum Einstein equations centered at a Schwarzschild metric, with suitably regular initial data, decay to a linearize…
Generative models create indistinguishable adversarial objects for object detection.
Accelerates optimization in asynchronous systems with sparse updates.
We study the perturbations of two classes of static black ellipsoid solutions of four dimensional vacuum Einstein equations. Such solutions are described by generic off--diagonal metrics which are generated by anholonomic transforms of diagonal metrics. The analysis is performed in the approximation of small eccentrici…
This paper generates natural-looking perturbations to fool classifiers.
PWGF escapes saddle points in nonconvex optimization.
Stochastic optimization algorithms with variance reduction have proven successful for minimizing large finite sums of functions. Unfortunately, these techniques are unable to deal with stochastic perturbations of input data, induced for example by data augmentation. In such cases, the objective is no longer a finite su…
Paper proposes EMO-based AE generation for black-box settings.
EdgeFool generates adversarial images to mislead classifiers.
Distributed descent-based methods are an essential toolset to solving optimization problems in multi-agent system scenarios. Here the agents seek to optimize a global objective function through mutual cooperation. Oftentimes, cooperation is achieved over a wireless communication network that is prone to delays and erro…
SGD converges with perturbed forward-backward passes, explained by geometric amplification.
New optimization method improves generalization across various tasks.
We explain how deformation theories of geometric objects such as complex structures, Poisson structures and holomorphic bundle structures lead to differential Gerstenhaber or Poisson algebras. We use homological perturbation theory to obtain algebra structures and some canonically defined deformations of s…
Sparse perturbations improve convergence in SZO methods for faster training.
Introduces a new geometric framework for non-perturbative BV-theory.
Proposes an efficient method for ordered counterfactual explanations.
Universal audio perturbations fool multiple classification models.
Item recommendation is a personalized ranking task. To this end, many recommender systems optimize models with pairwise ranking objectives, such as the Bayesian Personalized Ranking (BPR). Using matrix Factorization (MF) --- the most widely used model in recommendation --- as a demonstration, we show that optimizing it…
Recent work has documented the susceptibility of deep learning systems to adversarial examples, but most such attacks directly manipulate the digital input to a classifier. Although a smaller line of work considers physical adversarial attacks, in all cases these involve manipulating the object of interest, e.g., putti…
New online algorithms tackle dynamic decision-focused learning.
AWP improves robustness by flattening weight loss landscape.
SIGNNAP learns stable and identifiable node representations in GNNs against graph perturbations.
We present CROSSGRAD, a method to use multi-domain training data to learn a classifier that generalizes to new domains. CROSSGRAD does not need an adaptation phase via labeled or unlabeled data, or domain features in the new domain. Most existing domain adaptation methods attempt to erase domain signals using technique…
The paper analyzes and proposes methods for privately sharing individual privacy losses using per-instance differential privacy.
We use closed geodesics to construct and compute Bott-type Morse homology groups for the energy functional on the loop space of flat -dimensional tori, , and Bott-type Floer cohomology groups for their cotangent bundles equipped with the natural symplectic structure. Both objects are isomorpic to the singula…
ALPS improves neural network robustness and generalization.
The process of data mining with differential privacy produces results that are affected by two types of noise: sampling noise due to data collection and privacy noise that is designed to prevent the reconstruction of sensitive information. In this paper, we consider the problem of designing confidence intervals for the…
Exploring a new method to explain AI models in medical devices.
SGD generalization bounds derived from information theory.
This work improves PCA for robustness against adversarial perturbations.
SRO optimizes decisions against worst-case sampler induced by generative models.
This work learns visual representations for deformable objects using contrastive estimation.
Diffusion models optimize objectives similar to ELBO with Gaussian noise augmentation.
Hierarchical randomized smoothing improves model robustness for complex data.
Our main objective is to demonstrate how homological perturbation theory (HPT) results over the last 40 years immediately or with little extra work give some of the Koszul duality results that have appeared in the last decade. Higher homotopies typically arise when a huge object, e. g. a chain complex defining various …
In this paper, we consider efficient differentially private empirical risk minimization from the viewpoint of optimization algorithms. For strongly convex and smooth objectives, we prove that gradient descent with output perturbation not only achieves nearly optimal utility, but also significantly improves the running …
We are focusing on bound constrained global optimization problems, whose objective functions are computationally expensive black-box functions and have multiple local minima. The recently popular Metric Stochastic Response Surface (MSRS) algorithm proposed by \cite{Regis2007SRBF} based on adaptive or sequential learnin…
Introduces TPV to analyze model robustness without labels.
Stochastic zeroth-order (SZO), or gradient-free, optimization allows to optimize arbitrary functions by relying only on function evaluations under parameter perturbations, however, the iteration complexity of SZO methods suffers a factor proportional to the dimensionality of the perturbed function. We show that in scen…
ManiFlow models manifold data by optimizing NFs on perturbed data.