中文
相关论文

相关论文: Agnostic Learning under Targeted Poisoning: Optima…

200 篇论文

Consider the task of learning a hypothesis class $\mathcal{H}$ in the presence of an adversary that can replace up to an $\eta$ fraction of the examples in the training set with arbitrary adversarial examples. The adversary aims to fail the…

机器学习 · 计算机科学 2022-10-13 Steve Hanneke , Amin Karbasi , Mohammad Mahmoody , Idan Mehalel , Shay Moran

We construct algorithms with optimal error for learning with adversarial noise. The overarching theme of this work is that the use of \textsl{randomized} hypotheses can substantially improve upon the best error rates achievable with…

数据结构与算法 · 计算机科学 2026-04-06 Guy Blanc

Statistical learning theory and the Probably Approximately Correct (PAC) criterion are the common approach to mathematical learning theory. PAC is widely used to analyze learning problems and algorithms, and have been studied thoroughly.…

机器学习 · 计算机科学 2024-05-03 Adi Hendel , Meir Feder

Binary classification in the classic PAC model exhibits a curious phenomenon: Empirical Risk Minimization (ERM) learners are suboptimal in the realizable case yet optimal in the agnostic case. Roughly speaking, this owes itself to the fact…

机器学习 · 计算机科学 2025-12-22 Julian Asilis , Mikael Møller Høgsgaard , Grigoris Velegkas

A classical result in online learning characterizes the optimal mistake bound achievable by deterministic learners using the Littlestone dimension (Littlestone '88). We prove an analogous result for randomized learners: we show that the…

机器学习 · 计算机科学 2025-12-18 Yuval Filmus , Steve Hanneke , Idan Mehalel , Shay Moran

We study bandit best-arm identification with arbitrary and potentially adversarial rewards. A simple random uniform learner obtains the optimal rate of error in the adversarial scenario. However, this type of strategy is suboptimal when the…

机器学习 · 统计学 2026-04-17 Yasin Abbasi-Yadkori , Peter L. Bartlett , Victor Gabillon , Alan Malek , Michal Valko

In this paper, we study PAC learnability and certification of predictions under instance-targeted poisoning attacks, where the adversary who knows the test instance may change a fraction of the training set with the goal of fooling the…

机器学习 · 计算机科学 2021-08-10 Ji Gao , Amin Karbasi , Mohammad Mahmoody

Consider the domain of multiclass classification within the adversarial online setting. What is the price of relying on bandit feedback as opposed to full information? To what extent can an adaptive adversary amplify the loss compared to an…

机器学习 · 计算机科学 2024-02-13 Yuval Filmus , Steve Hanneke , Idan Mehalel , Shay Moran

We revisit the problem of private online learning, in which a learner receives a sequence of $T$ data points and has to respond at each time-step a hypothesis. It is required that the entire stream of output hypotheses should satisfy…

机器学习 · 计算机科学 2025-11-11 Bo Li , Wei Wang , Peng Ye

Online strategic classification studies settings in which agents strategically modify their features to obtain favorable predictions. For example, given a classifier that determines loan approval based on credit scores, applicants may open…

机器学习 · 计算机科学 2026-02-09 Chase Hutton , Adam Melrod , Han Shao

We establish a tight characterization of the worst-case rates for the excess risk of agnostic learning with sample compression schemes and for uniform convergence for agnostic sample compression schemes. In particular, we find that the…

机器学习 · 计算机科学 2018-05-22 Steve Hanneke , Aryeh Kontorovich

We study data poisoning attacks in the online setting where training items arrive sequentially, and the attacker may perturb the current item to manipulate online learning. Importantly, the attacker has no knowledge of future training items…

机器学习 · 计算机科学 2019-06-03 Xuezhou Zhang , Xiaojin Zhu , Laurent Lessard

We study the problem of robust learning under clean-label data-poisoning attacks, where the attacker injects (an arbitrary set of) correctly-labeled examples to the training set to fool the algorithm into making mistakes on specific test…

机器学习 · 计算机科学 2021-07-08 Avrim Blum , Steve Hanneke , Jian Qian , Han Shao

We study problem-dependent rates, i.e., generalization errors that scale near-optimally with the variance, the effective loss, or the gradient norms evaluated at the "best hypothesis." We introduce a principled framework dubbed "uniform…

机器学习 · 统计学 2020-12-25 Yunbei Xu , Assaf Zeevi

We study online learning in the adversarial injection model introduced by [Goel et al. 2017], where a stream of labeled examples is predominantly drawn i.i.d.\ from an unknown distribution $\mathcal{D}$, but may be interspersed with…

机器学习 · 计算机科学 2026-02-24 Ezra Edelman , Surbhi Goel

We study the problem of agnostically learning halfspaces which is defined by a fixed but unknown distribution $\mathcal{D}$ on $\mathbb{Q}^n\times \{\pm 1\}$. We define $\mathrm{Err}_{\mathrm{HALF}}(\mathcal{D})$ as the least error of a…

计算复杂性 · 计算机科学 2016-03-15 Amit Daniely

In this work, we initiate a formal study of probably approximately correct (PAC) learning under evasion attacks, where the adversary's goal is to \emph{misclassify} the adversarially perturbed sample point $\widetilde{x}$, i.e.,…

机器学习 · 计算机科学 2019-06-14 Dimitrios I. Diochnos , Saeed Mahloujifar , Mohammad Mahmoody

Data poisoning is a training-time attack that undermines the trustworthiness of learned models. In a targeted data poisoning attack, an adversary manipulates the training dataset to alter the classification of a targeted test point. Given…

机器学习 · 计算机科学 2025-11-18 Nakshatra Gupta , Sumanth Prabhu , Supratik Chakraborty , R Venkatesh

The current paper studies the problem of agnostic $Q$-learning with function approximation in deterministic systems where the optimal $Q$-function is approximable by a function in the class $\mathcal{F}$ with approximation error $\delta \ge…

机器学习 · 计算机科学 2020-02-18 Simon S. Du , Jason D. Lee , Gaurav Mahajan , Ruosong Wang

Multi-distribution learning extends agnostic Probably Approximately Correct (PAC) learning to the setting in which a family of $k$ distributions, $\{D_i\}_{i\in[k]}$, is considered and a classifier's performance is measured by its error…

机器学习 · 计算机科学 2025-06-24 Chicheng Zhang , Yihan Zhou
‹ 上一页 1 2 3 10 下一页 ›