Systematic Poisoning Attacks on and Defenses for Machine Learning in Healthcare

By Mehran Mozaffari Kermani, Susmita Sur-Kolay, Anand Raghunathan, Niraj Kumar Jha. 2015

347 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Safety, Robustness & Interpretability

How researchers see this paper

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

In Artificial Intelligence

In Safety, Robustness & Interpretability

Citations in the corpus

Towards the Science of Security and Privacy in Machine Learning (2016, methodology): “A similar attack approach has been applied in the context of healthcare [45].”
Neural Cleanse: Identifying and Mitigating Backdoor Attacks in Neural Networks (2019, background): “Defenses against poisoning attack mostly focus on sanitizing the training set and removing poisoned samples [32], [33], [34], [35], [36], [37].”
Mitigating Sybils in Federated Learning Poisoning (2018, background): “In a poisoning attack [6], [25], an adversary meticulously creates adversarial training examples and inserts them into the training data set of an attacked model.”
Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learning (2018, background): “In the security community, practical poisoning attacks have been demonstrated in worm signature generation [41], [44], spam filters [39], network traffic analysis systems for detection of DoS attacks [46], sentiment analysis on social networks [40], crowdsourcing [53], and health-care [37].”
When Does Machine Learning FAIL? Generalized Transferability for Evasion and Poisoning Attacks (2018, background): “Targeted poisoning attacks [34, 32, 11] require control of the labeling process.”
Adversarial Examples: Attacks and Defenses for Deep Learning (2017, background): “, training data poisoning) is another interesting topic and has been studied in [8], [16], [21], [98], [112], [142].”
Generative Poisoning Attack Method Against Neural Networks (2017, methodology): “A systematic, algorithm-independent poisoning attack approach was presented in [13].”
SoK: Security and Privacy in Machine Learning (2018, methodology): “A similar attack approach has been applied in the context of healthcare [38].”
Hardware Trojan Attacks on Neural Networks (2018, background): “However, due to the ubiquity and complexity of machine learning especially deep neural networks, it has been shown recently that these techniques are quite vulnerable to well-crafted attacks [8]–[11], which raised security concerns in the practical deployment of machine learning technologies.”
Latent Backdoor Attacks on Deep Neural Networks (2019, background): “Defenses against poisoning attacks mostly focus on sanitizing training data and removing poisoned samples [6, 15, 21, 36, 44, 47].”
Auror: defending against poisoning attacks in collaborative deep learning systems (2016, background): “Previous work has considered data sanitization to measure the negative impact of training dataset towards a trusted model [29, 32].”
Interpretable Machine Learning in Healthcare (2018, background): “critical use cases that include clinical decision making, there is some hesitation in the deployment of such models because the cost of model misclassification is potentially high [21].”
ABS: Scanning Neural Networks for Back-doors by Artificial Brain Stimulation (2019, background): “Many defense techniques have been proposed to defend against training data poisoning [16, 27, 44, 48].”
Mitigating Poisoning Attacks on Machine Learning Models: A Data Provenance Based Approach (2017, background): “Similarly, [12] present an attack that targets the labels input into the training system and a threshold based methodology to detect poison that relies on a Kappa statistic.”
Secure and Robust Machine Learning for Healthcare: A Survey (2020, neutral): “instance, systematic poisoning attacks against six conventional ML models that were developed for hypothyroid diagnosis are presented in [103], where the objective of the attacker was to”