Bagging Classifiers for Fighting Poisoning Attacks in Adversarial Classification Tasks

By Battista Biggio, Igino Corona, Giorgio Fumera, Giorgio Giacinto, Fabio Roli. 2011

136 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications, 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 Computer Networks and Communications

In Safety, Robustness & Interpretability

Citations in the corpus

Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning (2017, background): “, attack detection and removal) [94, 97, 99], and robust learning”
Mitigating Sybils in Federated Learning Poisoning (2018, background): “Another poisoning defense involves bagging classifiers [7] to mitigate the effects of outliers in the system.”
Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization (2017, neutral): “Besides the extension and evaluation of this poisoning attack strategy to different deep learning architectures and nonparametric models, further research avenues include: the investigation of the existence of universal perturbations (not dependent on the initial attack point) for poisoning samples ”
Attacks on Machine Learning: Lurking Danger for Accountability (2019, background): “TIR Poisoning Attack [60] X [14, 47] [10, 13, 14, 35, 39, 47] X [10, 14, 35, 36, 39, 47, 55, 67] X”
A Comprehensive Study of Security of Internet-of-Things (2017, unknown): “[147].”
A Survey on Security Threats and Defensive Techniques of Machine Learning: A Data Driven View (2018, methodology): “Hence, ensuring the purity of training data [76] and improving the robustness of learning algorithms [77]–[79] are two main countermeasures towards such adversary at the training phase.”
The relationship between trust in AI and trustworthy machine learning technologies (2019, unknown): “, in the training stage [26, 13, 11, 61, 85].”
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.”
Systematic Poisoning Attacks on and Defenses for Machine Learning in Healthcare (2015, unknown): “Machine-learning methods that account for data manipulation by adversaries (robust classifiers) have also been investigated [25], [26].”
Adversarial attacks against intrusion detection systems: Taxonomy, solutions and open issues (2013, methodology): “[103] propose the Reject On Negative Impact (RONI) technique: it is able to filter out 100% of poisoning attacks with 0% FP (poisoning fraction=5%) [11] show that weighted bagging can successfully handle 20% of poisoning attacks in the training set, without affecting IDS accuracy [131] show that rob”
Auror: defending against poisoning attacks in collaborative deep learning systems (2016, background): “Defenses against poisoning attacks are known in the area of collaborative recommendation systems using machine learning techniques [15,19,36].”
Machine learning for email spam filtering: review, approaches and open research problems (2019, methodology): “Biggio 8549 0 100 2005 Biggio et al [39]”
Curie: A method for protecting SVM Classifier from Poisoning Attack (2016, methodology): “Another method is to use multiple classifiers [4].”
Robust Linear Regression Against Training Data Poisoning (2017, background): “A number of heuristic techniques have also been proposed for poisoning attacks [3, 5, 25] for other problems, such as robust anomaly detection source identification.”
Robust High-Dimensional Linear Regression (2016, background): “A number of heuristic techniques have also been proposed for poisoning attacks [14, 15, 16] for other problems, such as robust anomaly detection source identification.”