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 19 of 19 for “"machine unlearning"”.
-
MACHINE UNLEARNING IN GRAPHS, RECOMMENDATIONS, AND LANGUAGE MODELS
Selective forgetting in machine learning has emerged as a transformative paradigm for addressing critical challenges related to data privacy, ethical compliance, and adaptability in dynamic environments. The rapid advancements in this field have intensified concerns over these issues, necessitating …
-
Machine Unlearning for Mobility Data: An Algorithmic Perspective
This work addresses machine unlearning for trajectory data, sequences of spatiotemporal points representing movement. Motivated by growing privacy concerns and regulations like GDPR and CCPA, which grant users the right to request deletion of their personal data from trained models (the right to be …
-
A Secure and Privacy-Preserving, Meta-Learning-Based Machine Unlearning Approach Using Redactable Blockchain under Intent-Based Networking Orchestration
… learned influence from AI models. Traditional unlearning methods that retrain models from scratch are computationally costly and unsuitable for infrastructures. To address this, this thesis proposes a secure and privacy-preserving machine unlearning framework that leverages meta-learning to …
-
Mašininio pamiršimo poreikio ir ateities perspektyvų analizė /
… benefits, and challenges of implementing machine unlearning technology based on the analysis of scientific literature and the conducted research, as well as to develop a recommendation-based structured theoretical framework for further research and practical application of the technology. …
-
Rethinking adversarial robustness in the context of the right to be forgotten
… which has motivated researchers to develop machine unlearning methods to unlearn a fraction of training data and its lineage. While existing machine unlearning methods prioritize the protection of individuals' private data, they overlook investigating the unlearned models' susceptibility to …
-
Adaptive and Responsible Machine Learning: From Incremental Learning to Unlearning and Safe Generation
Modern machine learning (ML) systems are increasingly required to operate in dynamic environments, where models must adapt to new information, remove obsolete or sensitive knowledge, and satisfy fairness and safety requirements. This thesis addresses these challenges by developing mechanisms for …
-
Theoretically-grounded efficient deep learning system design
… prior information from the training set, i.e., ''machine unlearning.'' This effort is particularly tailored for image-to-image generative models. By introducing a unified framework and a computationally-efficient algorithm, the proposed approach effectively removes information from the "forget" …
-
Statistical Learning via Stochastic Optimization under Data Privacy Considerations
… case. The second part concerns the problem of machine unlearning. The goal here is to efficiently update a trained model under requests to unlearn a data point in the training dataset. We delve into the problem for widely-studied classes of convex losses: smooth/non-smooth settings and …
-
Investigating and Evaluating Poison Unlearning under Imperfect Detection
Machine unlearning is an emerging field focused on removing the influence of specific data points from trained machine learning models. Early work in this field has primarily focused on privacy-oriented unlearning, often motivated by data protection regulations such as the GDPR. More recently, …
-
Towards smarter, safer, and swifter generative AI: adapting, guarding, and accelerating diffusion models
… Distillation (SFD), a swift, data-free machine unlearning method that fosters trustworthiness by erasing unsafe concepts through efficient fine-tuning rather than prohibitive re-training. Finally, to make models swifter, we present DLM-One, a score-distillation-based framework that …
-
Order-Leading Branch and Bound for Neural Network Verification
… errors across continual retraining, pruning, or unlearning throughout the lifecycle. This thesis addresses that gap by reconceiving deep learning model verification as a scalable, life-cycle-aware service rather than a one-off, model-centric exercise. The thesis is reasoning and opening the black …
-
Privacy-preserving computations and machine learning over decentralized data
In the current era, Machine Learning (ML) applications are ubiquitous in our daily lives, triggering major milestones in our interaction with technology. Artificial Intelligence (AI) and ML have reshaped our approach to productivity by enhancing a wide range of activities and contributing to …
-
Machine learning and unlearning for IoT anomaly detection
… IoT anomaly detection systems, we propose a new machine unlearning method called ViFLa, which groups training data based on estimated unlearning probability and treats each group as a virtual client in the federated learning framework. Since the virtual clients are physically in the same machine, …
-
Privacy Protection in Machine Learning: Methods for Structured and Unstructured Data
As machine learning models become increasingly integrated into data-driven decision-making, the protection of sensitive information throughout the model lifecycle is a paramount concern. As these models process and memorize sensitive, proprietary, or personal data, they risk leaking information …
-
PRIVACY-PRESERVING DATA GENERATION AND REMOVAL IN GENERATIVE MODELS
… vulnerabilities in existing fine-tuning-based unlearning methods through the Diffusion Model Relearning Attack (DiMRA), showing that removed content can be recovered. To address this, it proposes Diffusion Model Unlearning by Memorization (DiMUM). It also formalizes sample-level unlearning, …
-
New definitions, measurements, and algorithms for private data analysis
This thesis is about the theory of computing with data while protecting privacy. In this thesis, we study novel privacy settings and explore fundamental trade-offs between accuracy and information disclosure therein. The problems and solutions we consider are unified by a rigorous treatment of …
-
Understanding and Adapting Tree Ensembles: A Training Data Perspective
Despite the impressive success of deep-learning models on unstructured data (e.g., images, audio, text), tree-based ensembles such as random forests and gradient-boosted trees are hugely popular and remain the preferred choice for tabular or structured data, and are regularly used to win challenges …
-
Secure Machine Learning in Networking Systems
With the rapid integration of Machine Learning (ML) into networking systems, ensuring the security and trustworthiness of these intelligent frameworks has become a paramount concern. While ML offers unprecedented capabilities in spectrum management and collaborative learning, it also introduces …
-
Topics in efficient and privacy-preserving storage system design
Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2024-12-01