It begins with a boundary: A geometric view on probabilistically robust learning

• URL: http://arxiv.org/abs/2305.18779v3
• Date: Mon, 30 Sep 2024 14:07:43 GMT
• Title: It begins with a boundary: A geometric view on probabilistically robust learning
• Authors: Leon Bungert, Nicolás García Trillos, Matt Jacobs, Daniel McKenzie, Đorđe Nikolić, Qingsong Wang,
• Abstract summary: We take a fresh and geometric view on one such method -- Probabilistically Robust Learning (PRL) We prove existence of solutions to the original and modified problems using novel relaxation methods. We also clarify, through a suitable $Gamma$-convergence analysis, the way in which the original and modified PRL models interpolate between risk minimization and adversarial training.
• Score: 6.877576704011329
• License:
• Abstract: Although deep neural networks have achieved super-human performance on many classification tasks, they often exhibit a worrying lack of robustness towards adversarially generated examples. Thus, considerable effort has been invested into reformulating standard Risk Minimization (RM) into an adversarially robust framework. Recently, attention has shifted towards approaches which interpolate between the robustness offered by adversarial training and the higher clean accuracy and faster training times of RM. In this paper, we take a fresh and geometric view on one such method -- Probabilistically Robust Learning (PRL). We propose a mathematical framework for understanding PRL, which allows us to identify geometric pathologies in its original formulation and to introduce a family of probabilistic nonlocal perimeter functionals to rectify them. We prove existence of solutions to the original and modified problems using novel relaxation methods and also study properties, as well as local limits, of the introduced perimeters. We also clarify, through a suitable $\Gamma$-convergence analysis, the way in which the original and modified PRL models interpolate between risk minimization and adversarial training.

Related papers

• Regularization for Adversarial Robust Learning [18.46110328123008]
  We develop a novel approach to adversarial training that integrates $phi$-divergence regularization into the distributionally robust risk function. This regularization brings a notable improvement in computation compared with the original formulation. We validate our proposed method in supervised learning, reinforcement learning, and contextual learning and showcase its state-of-the-art performance against various adversarial attacks.
  arXiv  Detail & Related papers  (2024-08-19T03:15:41Z)
• Provable Risk-Sensitive Distributional Reinforcement Learning with General Function Approximation [54.61816424792866]
  We introduce a general framework on Risk-Sensitive Distributional Reinforcement Learning (RS-DisRL), with static Lipschitz Risk Measures (LRM) and general function approximation. We design two innovative meta-algorithms: textttRS-DisRL-M, a model-based strategy for model-based function approximation, and textttRS-DisRL-V, a model-free approach for general value function approximation.
  arXiv  Detail & Related papers  (2024-02-28T08:43:18Z)
• Provably Efficient Partially Observable Risk-Sensitive Reinforcement Learning with Hindsight Observation [35.278669159850146]
  We introduce a novel formulation that integrates hindsight observations into a Partially Observable Decision Process (POMDP) framework. We develop the first provably efficient RL algorithm tailored for this setting. These techniques are of particular interest to the theoretical study of reinforcement learning.
  arXiv  Detail & Related papers  (2024-02-28T08:24:06Z)
• Capsa: A Unified Framework for Quantifying Risk in Deep Neural Networks [142.67349734180445]
  Existing algorithms that provide risk-awareness to deep neural networks are complex and ad-hoc. Here we present capsa, a framework for extending models with risk-awareness.
  arXiv  Detail & Related papers  (2023-08-01T02:07:47Z)
• The Dynamics of Riemannian Robbins-Monro Algorithms [101.29301565229265]
  We propose a family of Riemannian algorithms generalizing and extending the seminal approximation framework of Robbins and Monro. Compared to their Euclidean counterparts, Riemannian algorithms are much less understood due to lack of a global linear structure on the manifold. We provide a general template of almost sure convergence results that mirrors and extends the existing theory for Euclidean Robbins-Monro schemes.
  arXiv  Detail & Related papers  (2022-06-14T12:30:11Z)
• The Geometry of Adversarial Training in Binary Classification [1.2891210250935143]
  We establish an equivalence between a family of adversarial training problems for non-parametric binary classification and a family of regularized risk minimization problems. The resulting regularized risk minimization problems admit exact convex relaxations of the type $L1+$ (nonlocal) $operatornameTV$.
  arXiv  Detail & Related papers  (2021-11-26T17:19:50Z)
• ROMAX: Certifiably Robust Deep Multiagent Reinforcement Learning via Convex Relaxation [32.091346776897744]
  Cyber-physical attacks can challenge the robustness of multiagent reinforcement learning. We propose a minimax MARL approach to infer the worst-case policy update of other agents.
  arXiv  Detail & Related papers  (2021-09-14T16:18:35Z)
• Minimum-Delay Adaptation in Non-Stationary Reinforcement Learning via Online High-Confidence Change-Point Detection [7.685002911021767]
  We introduce an algorithm that efficiently learns policies in non-stationary environments. It analyzes a possibly infinite stream of data and computes, in real-time, high-confidence change-point detection statistics. We show that (i) this algorithm minimizes the delay until unforeseen changes to a context are detected, thereby allowing for rapid responses.
  arXiv  Detail & Related papers  (2021-05-20T01:57:52Z)
• Attribute-Guided Adversarial Training for Robustness to Natural Perturbations [64.35805267250682]
  We propose an adversarial training approach which learns to generate new samples so as to maximize exposure of the classifier to the attributes-space. Our approach enables deep neural networks to be robust against a wide range of naturally occurring perturbations.
  arXiv  Detail & Related papers  (2020-12-03T10:17:30Z)
• A Hamiltonian Monte Carlo Method for Probabilistic Adversarial Attack and Learning [122.49765136434353]
  We present an effective method, called Hamiltonian Monte Carlo with Accumulated Momentum (HMCAM), aiming to generate a sequence of adversarial examples. We also propose a new generative method called Contrastive Adversarial Training (CAT), which approaches equilibrium distribution of adversarial examples. Both quantitative and qualitative analysis on several natural image datasets and practical systems have confirmed the superiority of the proposed algorithm.
  arXiv  Detail & Related papers  (2020-10-15T16:07:26Z)
• Reparameterized Variational Divergence Minimization for Stable Imitation [57.06909373038396]
  We study the extent to which variations in the choice of probabilistic divergence may yield more performant ILO algorithms. We contribute a re parameterization trick for adversarial imitation learning to alleviate the challenges of the promising $f$-divergence minimization framework. Empirically, we demonstrate that our design choices allow for ILO algorithms that outperform baseline approaches and more closely match expert performance in low-dimensional continuous-control tasks.
  arXiv  Detail & Related papers  (2020-06-18T19:04:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.