Defending Against Adversarial Denial-of-Service Data Poisoning Attacks

Abstract

Data poisoning is one of the most relevant security threats against machine learning and data-driven technologies. Since many applications rely on untrusted training data, an attacker can easily craft malicious samples and inject them into the training dataset to degrade the performance of machine learning models. As recent work has shown, such Denial-of-Service (DoS) data poisoning attacks are highly effective. To mitigate this threat, we propose a new approach of detecting DoS poisoned instances. In comparison to related work, we deviate from clustering and anomaly detection based approaches, which often suffer from the curse of dimensionality and arbitrary anomaly threshold selection. Rather, our defence is based on extracting information from the training data in such a generalized manner that we can identify poisoned samples based on the information present in the unpoisoned portion of the data. We evaluate our defence against two DoS poisoning attacks and seven datasets, and find that it reliably identifies poisoned instances. In comparison to related work, our defence improves false positive / false negative rates by at least 50%, often more.

Published at ACSAC DYNAMICS 2020

References in corpus (8)

• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Understanding Black-box Predictions via Influence Functions
• Poisoning Attacks against Support Vector Machines
• Achieving Human Parity in Conversational Speech Recognition
• An Empirical Evaluation of Deep Learning on Highway Driving
• Is feature selection secure against training data poisoning?
• Detection of Adversarial Training Examples in Poisoning Attacks through Anomaly Detection
• Fingerprint Classification Based on Depth Neural Network