Enabling Accurate And Private Machine Learning Systems
Download Enabling Accurate And Private Machine Learning Systems PDF/ePub or read online books in Mobi eBooks. Click Download or Read Online button to get Enabling Accurate And Private Machine Learning Systems book now. This website allows unlimited access to, at the time of writing, more than 1.5 million titles, including hundreds of thousands of titles in various foreign languages.
Enabling Accurate and Private Machine Learning Systems
Machine learning applications in fields where data is sensitive, such as healthcare and banking, face challenges due to the need to protect the privacy of participating users. Tools developed in the past decades that aim to address this challenge include differential privacy and federated learning. Yet maintaining performance while protecting sensitive data poses a trilemma between accuracy, privacy, and efficiency. In this thesis, we aim to address these fundamental challenges and take a step towards enabling machine learning under privacy and resource constraints. On the differential privacy front, we develop in Chapter 2 an algorithm that addresses efficiency and accuracy of differentially private empirical risk minimization. We provide a dimension independent excess risk bound and show the algorithm converges to this excess risk bound at the same rate as AdaGrad. In Chapter 3 we introduce an algorithm for differentially private Top-k selection, a problem that often arises as a building block of large-scale data analysis tasks like NLP and recommender systems. The algorithm samples from a distribution with exponentially large support only in polynomial time and space, and improves existing pure differential privacy methods. On the federated learning front, locality of data imposes various system design challenges due to resource constraints. In Chapter 4, we propose federated and differentially private algorithms for matrix factorization tasks that arise when training recommender systems in the settings where data is distributed across different silos (e.g., hospitals or banks). Chapter 5 introduces a client selection strategy that reduces communication in federated learning while maintaining accuracy of the model. Finally, in Chapter 6 we conclude by presenting F3AST, a novel algorithm that addresses user intermittency in federated learning under an unknown and time varying system configuration
Enabling Blockchain Technology for Secure Networking and Communications
In recent years, the surge of blockchain technology has been rising due to is proven reliability in ensuring secure and effective transactions, even between untrusted parties. Its application is broad and covers public and private domains varying from traditional communication networks to more modern networks like the internet of things and the internet of energy crossing fog and edge computing, among others. As technology matures and its standard use cases are established, there is a need to gather recent research that can shed light on several aspects and facts on the use of blockchain technology in different fields of interest. Enabling Blockchain Technology for Secure Networking and Communications consolidates the recent research initiatives directed towards exploiting the advantages of blockchain technology for benefiting several areas of applications that vary from security and robustness to scalability and privacy-preserving and more. The chapters explore the current applications of blockchain for networking and communications, the future potentials of blockchain technology, and some not-yet-prospected areas of research and its application. This book is ideal for practitioners, stakeholders, researchers, academicians, and students interested in the concepts of blockchain technology and the potential and pitfalls of its application in different utilization domains.
Sustainable IoT and Data Analytics Enabled Machine Learning Techniques and Applications
This book provides a structured presentation of machine learning related to vision, speech, and natural language processing. It addresses the tools, techniques, and challenges of machine learning algorithm implementation, computation time, and the complexity of reasoning and modeling of different types of data. The book covers diverse topics such as semantic image segmentation, deep visual residual abstraction, brain–computer interfaces, natural language processing, traffic and signaling, driverless driving, and radiology. The majority of smart applications have a need for a sustainable Internet of things (IoT) and artificial intelligence. Active research trends and future directions of machine learning under big data analytics are also discussed. Machine learning is a class of artificial neural networks that have become dominant in various computer vision tasks, attracting interest across a variety of domains as they are a type of deep neural networks efficient in extracting meaningful information from visual imagery.