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 20 of 30 for “"model security"”.
-
Trustworthy AI in the Modern Era: Theories and Applications
… interconnected critical components of modern AI: model security, fairness, explainability, and data privacy. For model security, I aim to ensure that the model integrity and behavior are not compromised against potential malicious attacks, especially model backdoor and stealing attacks. I propose …
-
IntruSense: An Enhanced Physical Security System Using UWB
Physical security, being a multi-billion dollar industry has now transformed into an IoT based service industry. Traditional security systems include video sensing through cameras, laser sensors, security guards or motion detectors, all of which are either expensive, unreliable, consume lots of …
-
Bayesian Option Pricing with Asymmetric Variance Processes
… negative skewness and excess kurtosis. GARCH models can be successfully used to model security returns. This thesis utilizes NGARCH and Normal Mixture NGARCH models to price options. The Gibbs sampling method is explained and implemented for parametric inference, and Bayesian inference results …
-
A study on the decision factors in the delay of the ROK-U.S. wartime OPCON transfer: focusing on the domestic determinants
… first applied Graham Allison’s Rational Actor Model (security); second, it applied domestic politics from Allison’s bureaucratic politics model; and third, to fill a gap between security (RAM) and domestic politics, this thesis used Robert Putnam’s Two-Level Game theory to clarify …
-
Building in web application security at the requirements stage : a tool for visualizing and evaluating security trade-offs : a thesis presented in partial fulfilment of the requirements for the degree of Master of Information Science in Information Systems at Massey University, Albany, New Zealand
One dimension of Internet security is web application security. The purpose of this Design-science study was to design, build and evaluate a computer-based tool to support security vulnerability and risk assessment in the early stages of web application design. The tool facilitates risk assessment …
-
Towards Reliable and Trustworthy Deep Learning through Explainability and Interpretability
… reliability and trustworthiness of deep learning models. First, we develop new post-hoc explanation methods that improve feature attribution and concept-based explanations. These methods provide faithful decision cues by modeling meaningful feature interactions and extracting faithful coherent …
-
Optimal Sizing and Location of Static and Dynamic Reactive Power Compensation
… constraints are the key of various optimization models, identified as Optimal Power Flow (OPF) model, Security Constrained OPF (SCOPF) model, and Voltage Stability Constrained OPF model (VSCOPF).</p> <p>First, this work investigates the economic benefits from local reactive power compensation …
-
Evolving Threats and Defenses in Machine Learning: Focus on Model Inversion and Beyond
Machine learning (ML) models are increasingly integrated into critical real-world applications, raising concerns about security, privacy, and trustworthiness. Among various emerging threats, model inversion (MI) attacks stand out due to their potential to compromise the confidentiality of training …
-
Understanding and Extending Backdoor Attacks on Deep Neural Networks: From Behavioural and Mechanistic Analysis to Concept-Level Triggers
Backdoor attacks pose a serious security threat to deep neural networks by implanting hidden attacker-specified behaviours that are activated only when a trigger is present. Although many attacks and defenses have been evaluated using aggregate metrics such as clean accuracy, attack success rate, …
-
Visible Light Communication Cyber Security Vulnerabilities For Indoor And Outdoor Vehicle-To-Vehicle Communication
… than 5 billion users globally. As a result, the security and privacy of 5G wireless networks is an essential problem as those modern applications are in people's life everywhere. VLC security is as one of the core physical-layer security (PLS) solutions for 5G networks. Due to the fact that light …
-
An Interaction-based Software-Defined Security Model and Platform to secure cloud resources
… is in finding effective mechanisms for isolating security issues in cloud infrastructure. Isolation implies creating security boundaries for protecting cloud assets at different levels of a cloud security architecture. Building security boundaries is critical not only for recognizing security …
-
From Image Enhancement to Model Protection Integrating Generative AI and Secure Learning in Computer Vision
… key areas: image translation, denoising, and model security. The research encompasses novel methodologies and models that significantly advance existing techniques. This dissertation will not only provide valuable contributions to the academic community but also hold significant potential for …
-
Remedies for challenges in deep learning-based quantitative ultrasound imaging
Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2024-09-16 without embargo terms
-
Secure Distance Bounding
… through a number of application scenarios and model security against distance fraud (DF), Man-in-the-Middle (MiM), and collusion fraud (CF) attacks. We prove impossibility of security against these attacks without making physical assumptions. We propose approaches to the construction of secure …
-
Evaluating & engineering
… they rarely (a) have a comprehensive view of the security-related decisions that should be made, or (b) know how these decisions precisely affect the implementation. A vital decision is the selection of tools and methods that can best support a particular situation in order to shield an …
-
Stochastic Tools for Network Security: Anonymity Protocol Analysis and Network Intrusion Detection
… and the sharp increase of network crime, network security has become very important and received a lot of attention. In this dissertation, we model security issues as stochastic systems. This allows us to find weaknesses in existing security systems and propose new solutions. Exploring the …
-
Machine learning techniques for the preservation of data privacy
… information which cannot be shared with model owners due to privacy concerns. For example, healthcare providers may be bound by patient privacy laws. Without large publicly available datasets, useful models become impossible to train. Therefore, ML methods that preserve the privacy of …
-
A Novel Human Visual Psychophysics Based Approach to Distinguish Between Human Users and Computer Robots
… need for developing a secure and robust CAPTCHA model. Security and usability are the two fundamental issues that pose a trade-off in the design of a CAPTCHA. If a CAPTCHA model were too difficult for human users to solve, it would affect its usability, but making it easy would risk its security. …
-
SPARK: secure, privacy-aware, and robust algorithms for high-accuracy mobile indoor localization using machine learning
… fingerprints are collected and trained on the ML model, and (2) the online phase, where new Wi-Fi fingerprints from user devices are processed through the pre-trained ML model to predict the user's locations in real-time. This approach offers resilience against common environmental disturbances, …
-
The Chain-Link Fence Model: A Framework for Creating Security Procedures
Information technology security professionals are facing an ever growing threat to the networks that they defend. The process for creating procedures to help stem this threat is very difficult for security professionals. The Chain-Link Fence Model helps security professionals by guiding them …
Page 1 of 2