Guardado en:
Detalles Bibliográficos
Autores principales: Bhat, Sunay, Jiang, Jeffrey, Pooladzandi, Omead, Branch, Alexander, Pottie, Gregory
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:https://arxiv.org/abs/2405.18627
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929370952630272
author Bhat, Sunay
Jiang, Jeffrey
Pooladzandi, Omead
Branch, Alexander
Pottie, Gregory
author_facet Bhat, Sunay
Jiang, Jeffrey
Pooladzandi, Omead
Branch, Alexander
Pottie, Gregory
contents Train-time data poisoning attacks threaten machine learning models by introducing adversarial examples during training, leading to misclassification. Current defense methods often reduce generalization performance, are attack-specific, and impose significant training overhead. To address this, we introduce a set of universal data purification methods using a stochastic transform, $Ψ(x)$, realized via iterative Langevin dynamics of Energy-Based Models (EBMs), Denoising Diffusion Probabilistic Models (DDPMs), or both. These approaches purify poisoned data with minimal impact on classifier generalization. Our specially trained EBMs and DDPMs provide state-of-the-art defense against various attacks (including Narcissus, Bullseye Polytope, Gradient Matching) on CIFAR-10, Tiny-ImageNet, and CINIC-10, without needing attack or classifier-specific information. We discuss performance trade-offs and show that our methods remain highly effective even with poisoned or distributionally shifted generative model training data.
format Preprint
id arxiv_https___arxiv_org_abs_2405_18627
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PureGen: Universal Data Purification for Train-Time Poison Defense via Generative Model Dynamics
Bhat, Sunay
Jiang, Jeffrey
Pooladzandi, Omead
Branch, Alexander
Pottie, Gregory
Machine Learning
Artificial Intelligence
Cryptography and Security
Train-time data poisoning attacks threaten machine learning models by introducing adversarial examples during training, leading to misclassification. Current defense methods often reduce generalization performance, are attack-specific, and impose significant training overhead. To address this, we introduce a set of universal data purification methods using a stochastic transform, $Ψ(x)$, realized via iterative Langevin dynamics of Energy-Based Models (EBMs), Denoising Diffusion Probabilistic Models (DDPMs), or both. These approaches purify poisoned data with minimal impact on classifier generalization. Our specially trained EBMs and DDPMs provide state-of-the-art defense against various attacks (including Narcissus, Bullseye Polytope, Gradient Matching) on CIFAR-10, Tiny-ImageNet, and CINIC-10, without needing attack or classifier-specific information. We discuss performance trade-offs and show that our methods remain highly effective even with poisoned or distributionally shifted generative model training data.
title PureGen: Universal Data Purification for Train-Time Poison Defense via Generative Model Dynamics
topic Machine Learning
Artificial Intelligence
Cryptography and Security
url https://arxiv.org/abs/2405.18627