Read more
This book surveys state-of-the-art research on adversarial robustness of quantum machine learning algorithms. Despite their high efficiency and accuracy, classical ML and AI algorithms can be easily fooled by an adversary through manipulation or spoofing of data (also known as adversarial attacks), which poses serious security ramifications. On the other hand, the integration of quantum computing in ML and AI is progressing rapidly to create new quantum ML/AI models which are designed to fundamentally exploit quantum mechanical properties to gain advantages in aspects such as training speed or feature extraction accuracy. This raises the important question of whether quantum AI algorithms are as vulnerable as classical AI models. Recent work has shown that quantum AI algorithms are remarkably robust against adversarial attacks. This offers a unique opportunity to leverage quantum computing, specifically its unique properties like superposition and entanglement, to develop highly resistant quantum AI systems. This shift is crucial for enhancing the safety and reliability of AI in security-sensitive applications. This book provides a comprehensive overview of the research in the emerging field of quantum adversarial AI, presenting seminal work from world-leading quantum AI experts on quantum AI and its benchmarking against adversarial attacks. It provides an essential reference for graduate students and industry experts who are interested in quantum AI for security-sensitive autonomous systems.
List of contents
Fundamentals of Quantum Machine Learning and Robustness.- Adversarial Robustness in Quantum Machine Learning.- Adversarial Attack Transferability of Quantum and Classical Classifiers.- Fundamental questions on robustness and accuracy for classical and quantum learning algorithms.- Adversarial Threats in Quantum Machine Learning: A Survey of Attacks and Defenses.
About the author
Professor Muhammad Usman is Head of Quantum Systems and Principal Staff Member at CSIRO’s Data61 which is Australia National Research Organisation. He has over 15 years of research and teaching experience in the field of quantum computing with a track-record of over 120 research papers in high-impact international journals. At CSIRO, Professor Usman is leading a team of over 20 researchers working at the forefront of quantum algorithms, quantum software engineering, and quantum security. He is a fellow of the Australian Institute of Physics and serves on the executive editorial boards of two international journals (Nature Scientific Reports and IOP Nano Futures), a committee member of Standards Australia to help in standardisation of quantum technologies and have academic affiliations at the University of Melbourne and RMIT University. Professor Usman is the chair of organising committee of international conference on Quantum Techniques in Machine Learning 2024 (now serves on the Steering Committee), has delivered numerous invited talks in international conferences and has been invited on several panel discussions at national and international meetings. He was received the State of Victoria iAward 2024, Innovative of the Year 2023 Award by Defence Industry, Winner of the Australian Army Quantum Technology Challenge in three years in a row (2021, 2022 and 2023), Rising Stars in Computational Materials Science by Elsevier in 2020, and Dean’s Award for Excellence in Research (Early Career) at the University of Melbourne in 2019. Professor Usman is a recipient of prestigious international research fellowships from Fulbright USA (20005-2010) and DAAD Germany in 2010. Professor Usman is a passionate quantum educator and has been promoting quantum education among school children as part of the CSIRO’s STEM Scientists in Schools program.
Summary
This book surveys state-of-the-art research on adversarial robustness of quantum machine learning algorithms. Despite their high efficiency and accuracy, classical ML and AI algorithms can be easily fooled by an adversary through manipulation or spoofing of data (also known as adversarial attacks), which poses serious security ramifications. On the other hand, the integration of quantum computing in ML and AI is progressing rapidly to create new quantum ML/AI models which are designed to fundamentally exploit quantum mechanical properties to gain advantages in aspects such as training speed or feature extraction accuracy. This raises the important question of whether quantum AI algorithms are as vulnerable as classical AI models. Recent work has shown that quantum AI algorithms are remarkably robust against adversarial attacks. This offers a unique opportunity to leverage quantum computing, specifically its unique properties like superposition and entanglement, to develop highly resistant quantum AI systems. This shift is crucial for enhancing the safety and reliability of AI in security-sensitive applications. This book provides a comprehensive overview of the research in the emerging field of quantum adversarial AI, presenting seminal work from world-leading quantum AI experts on quantum AI and its benchmarking against adversarial attacks. It provides an essential reference for graduate students and industry experts who are interested in quantum AI for security-sensitive autonomous systems.
