Mitigating Poisoning Attacks on Machine Learning Models: A Data Provenance Based Approach

By Nathalie Baracaldo, Bryant Chen, Heiko Ludwig, Jaehoon Amir Safavi. 2017

125 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability, Privacy & Security

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

In Privacy & Security

Citations in the corpus

Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering (2018, methodology): “Defenses : General defenses against poisoning attacks on supervised learning methods were proposed by Nelson et al. (2009) and Baracaldo et al. (2017).”
Mitigating Sybils in Federated Learning Poisoning (2018, background): “[4] employs data provenance to prevent poisoning attacks.”
Biscotti: A Ledger for Private and Secure Peer-to-Peer Machine Learning (2018, methodology): “[7] employ data provenance as a measure against poisoning attacks by tracking the history of the training data and removing information from anomalous sources.”
Security and Privacy Challenges of Large Language Models: A Survey (2024, unknown): “Clear principles of using these LLMs and/or other AI tools in scientific explorations should be established to ensure transparency, fairness, and trustworthiness [211].”
Demon in the Variant: Statistical Analysis of DNNs for Robust Backdoor Contamination Detection (2019, background): “[2] proposed two general protections against backdoor attack.”
Trustworthy AI: From Principles to Practices (2021, background): “Moreover, recent research has shown that data provenance can be used to mitigate data poisoning [33], thus enhancing the robustness and security of an AI system.”
Poisoning Attack in Federated Learning using Generative Adversarial Nets (2019, methodology): “[23], are not suitable for federated learning because they all require the detector to access the participants’ training data and training model, which is contradicted with the design idea of federated learning.”
Biscotti: A Blockchain System for Private and Secure Federated Learning (2021, background): “Finally, Baracaldo et al. [80] employ data provenance as a measure against poisoning attacks by tracking the history of the training data and removing information from anomalous sources.”
Adversarial Robustness Toolbox v0.2.2 (2018, background): “The defences parameter is either a string or a list of strings selecting the defences to be applied from the following: – featsqueeze[1-8]? for feature squeezing, where the digit at the end indicates the bit depth (see Section 6.”
Adversarial Robustness Toolbox v1.0.0 (2018, background): “Defences for this threat aim to detect and ﬁlter malicious training data [5, 4, 46].”
The Limitations of Federated Learning in Sybil Settings (2020, background): “In the general ML setting, defenses rely on access to the training data [4] or access to the training process itself [7, 25, 45] to defend against adversaries.”