Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 12 of 12 for “"Poisoning Attacks"”.
-
Game of threads: Enabling asynchronous poisoning attacks
… of these codes by introducing asynchronous poisoning attacks. Our attack influences training outcome---e.g., degrades accuracy or biases the model towards an adversary-specified label---purely by scheduling asynchronous training threads in a malicious fashion. Since thread scheduling is …
-
A defensive strategy for detecting targeted adversarial poisoning attacks in machine learning trained malware detection models
… are susceptible to a variety of adversarial attacks, which allow an attacker to manipulate the machine learning model into performing an unwanted action, such as misclassifying data into the attackers desired class, or reducing the overall efficacy of the ML model. One current research area …
-
Provable stability defenses for targeted data poisoning
… data, these systems may be susceptible to data poisoning attacks where malicious users inject false training data in order to influence the learned model. While recent work has focused primarily on the untargeted case, where the attacker's goal is to increase overall error, much less is …
-
Towards machine learning models robust to adversarial examples and backdoor attacks
… systems: adversarial examples and backdoor data poisoning attacks. Specifically, in the first part of the thesis, we build a methodology for defending against adversarial examples that is the first one to provide non-trivial adversarial robustness against an adaptive adversary. In the second …
-
Deep Neural Network for Anomaly Detection
… (CPSs) services create new surfaces for cyberattacks. To safeguard these CPSs, anomaly detection (AD) that detects potential attacks/adversarial behaviors plays a pivotal role. This thesis aims to design novel deep neural models to handle four challenges of the AD problem to deal with …
-
Metagradient Descent: Differentiating Large-Scale Training
… methods, outperform accuracy-degrading data poisoning attacks by an order of magnitude, and automatically find competitive learning rate schedules.
-
Towards Secure and Resilient Machine Learning Systems
… them to adversarial threats. These include poisoning attacks in the training phase and evasion attacks during inference, both of which compromise model reliability and accuracy. To enhance the robustness of ML models, this dissertation presents a series of studies that (1) strengthen the …
-
Protecting vehicles from cyberattacks: context aware AI-based intrusion detection for vehicle CAN bus security.
… measures, rendering it susceptible to cyberattacks, endangering passenger safety. Although artificial intelligence (AI)-based intrusion detection systems (IDSs) can detect these attacks, achieving higher detection rates in near-real-time poses challenges. This research aims to enhance …
-
Building trustworthy machine learning systems in adversarial environments
… model. These adversarial scenarios, known as poisoning attack, adversarial example attack, and inference attack, have demonstrated that security, privacy, and robustness have become more important than ever for AI to gain wider adoption and societal trust. To address these problems, we …
-
Principled approaches to robust machine learning and beyond
… for robust PCA, new defenses for data poisoning attacks for stochastic optimization, and new defenses for watermarking attacks on deep nets. In all of these tasks, we demonstrate on both synthetic and real data sets that our performance is substantially better than the state-of-the-art, …
-
An integrated security Protocol communication scheme for Internet of Things using the Locator/ID Separation Protocol Network
… stage is vulnerable to masquerading and content poisoning attacks. Secondly, the addresses resolving stage, in the Locator/ID Separation Protocol the Map Server (MS) accepts Map-Request from Ingress Tunnel Routers and Egress Tunnel Routers. These routers in trun look up the database and return …
-
Information-fusion-based methods to improve the detection of advanced cyber threats
… and specification-based detection of attacks to provide alerts and improve visibility into the target network. We develop techniques that apply to general networked systems. However, to make the discussion concrete and reason about our design decisions, we have adopted two types of …