arXiv research

LAZO reduces query complexity and variance in ZO methods.

Zeroth-order optimization is an important research topic in machine learning. In recent years, it has become a key tool in black-box adversarial attack to neural network based image classifiers. However, existing zeroth-order optimization algorithms rarely extract second-order information of the model function. In this…

Paper proposes ZO-NGD for more efficient black-box attacks.

Zeroth-order (a.k.a, derivative-free) methods are a class of effective optimization methods for solving complex machine learning problems, where gradients of the objective functions are not available or computationally prohibitive. Recently, although many zeroth-order methods have been developed, these approaches still…

Improved zeroth-order algorithms for nonconvex optimization with reduced complexity and improved performance.

New method uses zeroth-order queries to approximate proximal sampling efficiently.

We propose a method for zeroth order stochastic convex optimization that attains the suboptimality rate of $\tilde{\mathcal{O}}(n^{7}T^{-1/2})$ after $T$ queries for a convex bounded function $f:{\mathbb R}^n\to{\mathbb R}$. The method is based on a random walk (the \emph{Ball Walk}) on the epigraph of the function. Th…

Paper proposes estimating gradients for zeroth-order nonconvex optimization.

We consider the problem of global optimization of an unknown non-convex smooth function with zeroth-order feedback. In this setup, an algorithm is allowed to adaptively query the underlying function at different locations and receives noisy evaluations of function values at the queried points (i.e. the algorithm has ac…

We consider the problem of optimizing a high-dimensional convex function using stochastic zeroth-order queries. Under sparsity assumptions on the gradients or function values, we present two algorithms: a successive component/feature selection algorithm and a noisy mirror descent algorithm using Lasso gradient estimate…

New algorithm reduces dimensionality in stochastic optimization.

Algorithm optimizes non-convex functions using dueling comparisons.

Paper proposes algorithms for solving nonconvex-nonconcave problems with complexity guarantees.

Recent studies have shown that adversarial examples in state-of-the-art image classifiers trained by deep neural networks (DNN) can be easily generated when the target model is transparent to an attacker, known as the white-box setting. However, when attacking a deployed machine learning service, one can only acquire t…

This paper analyzes and guarantees convergence of prior-guided ZO algorithms.

Simplifies noisy convex optimization with a new algorithm.

A new method for optimizing functions without gradients, improving efficiency and convergence.

Paper tackles constrained robust optimization without gradients.

A new one-point feedback scheme improves ZO algorithms for black-box optimization.

A new algorithm for optimizing huge-scale black-box problems with reduced memory usage.

MAZE attacks a model without data, using synthetic inputs.

Paper learns to optimize without gradients for better performance.

As application demands for zeroth-order (gradient-free) optimization accelerate, the need for variance reduced and faster converging approaches is also intensifying. This paper addresses these challenges by presenting: a) a comprehensive theoretical analysis of variance reduced zeroth-order (ZO) optimization, b) a nove…

Develops a functional generalization of Eldan's stochastic localization for optimization and privacy.

Paper tackles gradient-free minimax optimization with variance reduction for faster convergence.

We study the most practical problem setup for evaluating adversarial robustness of a machine learning system with limited access: the hard-label black-box attack setting for generating adversarial examples, where limited model queries are allowed and only the decision is provided to a queried data input. Several algori…

New analysis for black-box learning without gradients, improving generalization bounds.

Study on collaborative vs. non-collaborative online and bandit convex optimization.

Zeroth-order optimization methods lack inherent privacy guarantees.

Paper develops methods to optimize policies directly from human feedback without reward inference.

Derivative-free optimization has become an important technique used in machine learning for optimizing black-box models. To conduct updates without explicitly computing gradient, most current approaches iteratively sample a random search direction from Gaussian distribution and compute the estimated gradient along that…

In this paper, we propose and analyze zeroth-order stochastic approximation algorithms for nonconvex and convex optimization, with a focus on addressing constrained optimization, high-dimensional setting and saddle-point avoiding. To handle constrained optimization, we first propose generalizations of the conditional g…

New method for zeroth-order stochastic gradient algorithms provides confidence intervals.

Zeroth-order methods favor flat minima in machine learning.

Improves zeroth-order optimization for private machine learning with public data.

We study the problem of attacking a machine learning model in the hard-label black-box setting, where no model information is revealed except that the attacker can make queries to probe the corresponding hard-label decisions. This is a very challenging problem since the direct extension of state-of-the-art white-box at…

New optimization method improves generalization across various tasks.

Paper tackles sampling from non-log-concave distributions using denoising diffusion.

Two algorithms solve nonconvex minimax problems with linear constraints, achieving complexity guarantees.

Improved zeroth-order algorithms tackle nonconvex minimax problems with reduced complexity.

ConMeZO speeds up zeroth-order optimization for large language models.

RayS attack improves hard-label adversarial attacks by reducing query complexity and identifying false robust models.

A new method for MARL with partial observations reduces communication overhead.

Proximal gradient method has been playing an important role to solve many machine learning tasks, especially for the nonsmooth problems. However, in some machine learning problems such as the bandit model and the black-box learning problem, proximal gradient method could fail because the explicit gradients of these pro…

This paper reviews zeroth-order optimization in signal processing and machine learning.

Book covers tools for zeroth-order convex optimisation.

New algorithm tackles risk-aware learning problems efficiently.

Discretizations of Langevin diffusions provide a powerful method for sampling and Bayesian inference. However, such discretizations require evaluation of the gradient of the potential function. In several real-world scenarios, obtaining gradient evaluations might either be computationally expensive, or simply impossibl…