Statistical learning theory provides the mathematical foundations for understanding generalization from finite data. As quantum computing and quantum machine learning continue to develop, there is increasing interest in whether analogous principles can be formulated for learning settings involving quantum data, quantum models, or quantum computation. This paper offers an introductory perspective on quantum statistical learning theory. We begin by recalling the basic ideas of classical statistical learning theory, including empirical risk minimization, hypothesis classes, generalization, and sample complexity, and then discuss how these concepts must be adapted in the quantum setting. In particular, we distinguish between learning with classical data using quantum algorithms, learning with quantum models, and learning directly from quantum data. We then outline the central learning-theoretic objects in the quantum setting, including quantum hypotheses, measurement-based predictions, population and empirical risk, shot complexity, and information-extraction limitations. A simple illustrative example is used to highlight how finite measurements and quantum uncertainty affect learning performance. We conclude with a discussion of open problems and directions toward a more unified theory of generalization and learnability in quantum learning.
Statistical learning theory provides the mathematical foundations for understanding generalization from finite data. As quantum computing and quantum machine learning continue to develop, there is increasing interest in whether analogous principles can be formulated for learning settings involving quantum data, quantum models, or quantum computation. This paper offers an introductory perspective on quantum statistical learning theory. We begin by recalling the basic ideas of classical statistical learning theory, including empirical risk minimization, hypothesis classes, generalization, and sample complexity, and then discuss how these concepts must be adapted in the quantum setting. In particular, we distinguish between learning with classical data using quantum algorithms, learning with quantum models, and learning directly from quantum data. We then outline the central learning-theoretic objects in the quantum setting, including quantum hypotheses, measurement-based predictions, population and empirical risk, shot complexity, and information-extraction limitations. A simple illustrative example is used to highlight how finite measurements and quantum uncertainty affect learning performance. We conclude with a discussion of open problems and directions toward a more unified theory of generalization and learnability in quantum learning.
Abstract Learning problems involve settings in which an algorithm has to make decisions based on data, and possibly side information such as expert knowledge. This study has two main goals. First, it reviews and generalizes different results on the data and model complexity of quantum learning, where the data and/or the algorithm can be quantum, focusing on information‐theoretic techniques. Second, it introduces the notion of copy complexity, which quantifies the number of copies of a quantum state required to achieve a target accuracy level. Copy complexity arises from the destructive nature of quantum measurements, which irreversibly alter the state to be processed, limiting the information that can be extracted about quantum data. As a result, empirical risk minimization is generally inapplicable. The paper presents novel results on the copy complexity for both training and testing. To make the paper self‐contained and approachable by different research communities, an extensive background material is provided on classical results from statistical learning theory, as well as on the distinguishability of quantum states. Throughout, the differences between quantum and classical learning are highlighted by addressing both supervised and unsupervised learning, and extensive pointers are provided to the literature.
Over decades traditional information theory of source and channel coding advances toward learning and effective extraction of information from data. We propose to go one step further and offer a theoretical foundation for learning classical patterns from quantum data. However, there are several roadblocks to lay the groundwork for such a generalization. First, classical data must be replaced by a density operator over a Hilbert space. Hence, deviated from problems such as state tomography, our samples are i.i.d density operators. The second challenge is even more profound since we must realize that our only interaction with a quantum state is through a measurement which - due to no-cloning quantum postulate - loses information after measuring it. With this in mind, we present a quantum counterpart of the well-known probably approximately correct (PAC) framework. Based on that, we propose a quantum analogous of the Empirical Risk Minimization (ERM) algorithm for learning measurement hypothesis classes. Then, we establish upper bounds on the quantum sample complexity quantum concept classes.
We propose a learning model called the quantum statistical learning QSQ model, which extends the SQ learning model introduced by Kearns to the quantum setting. Our model can be also seen as a restriction of the quantum PAC learning model: here, the learner does not have direct access to quantum examples, but can only obtain estimates of measurement statistics on them. Theoretically, this model provides a simple yet expressive setting to explore the power of quantum examples in machine learning. From a practical perspective, since simpler operations are required, learning algorithms in the QSQ model are more feasible for implementation on near-term quantum devices. We prove a number of results about the QSQ learning model. We first show that parity functions, (log n)-juntas and polynomial-sized DNF formulas are efficiently learnable in the QSQ model, in contrast to the classical setting where these problems are provably hard. This implies that many of the advantages of quantum PAC learning can be realized even in the more restricted quantum SQ learning model. It is well-known that weak statistical query dimension, denoted by WSQDIM(C), characterizes the complexity of learning a concept class C in the classical SQ model. We show that log(WSQDIM(C)) is a lower bound on the complexity of QSQ learning, and furthermore it is tight for certain concept classes C. Additionally, we show that this quantity provides strong lower bounds for the small-bias quantum communication model under product distributions. Finally, we introduce the notion of private quantum PAC learning, in which a quantum PAC learner is required to be differentially private. We show that learnability in the QSQ model implies learnability in the quantum private PAC model. Additionally, we show that in the private PAC learning setting, the classical and quantum sample complexities are equal, up to constant factors.
The intersection of machine learning and quantum computing has been one of the greatest scientific projects in current time. With quantum computing in the offing, the assurance of using quantum mechanical effects in computational learning to transform the world has never been closer. The publication of this book comes at a critical time, when theoretical models are shouldering maturity with capabilities of experimentation, that has brought about great possibilities to implement computational benefits that quantum systems can provide in performing machine learning tasks. Quantum machine learning is one field which has rapidly changed in the last ten years. What was originally viewed as mere theoretical conjecture has now evolved into a living research community which includes rigorous mathematical modeling, novel algorithm design, and more and more complex experimental realisations on the near-term quantum computers. Such change can be characterised by the increasing sophistication of our theory in as well as the revolutionary advances in quantum computing technology, and by the noisy intermediate-scale quantum (NISQ) era. Chapters in this book chronologically discuss the complex nature of quantum machine learning, including theoretical principles and challenges of practice. All the chapters cover a very important aspect in this growing area leaving the reader not only with an in-depth analysis but also a wide enough point of view to traverse this somewhat complicated inter-disciplinary area. The book is organized in such a way that that it takes the reader through more advanced parts of the quantum machine learning theory and practice. We start by explaining quantum feature maps, and kernel-type methods and laying the mathematical background on which quantum-enhanced learning is based. There we discuss computational complexity, trainability issues, neural network architectures, scaling of resources, sample complexity limits and the issue critical of quantum data encoding. This book has a variety of audience. The works will also offer exhaustive proposals of key concepts, methodologies and open problems to graduate students and researchers venturing into the field. Individuals who operate quantum computing platforms will have an understanding of the options available when it comes to designing an algorithm and the resources needed, as well as the expectations of its performance. The systematic approach of tackling complexity theorist-foundations and rigorous mathematical analyses will be appreciated by the theorists. Quantum machine learning is interdisciplinary and requires the services and knowledge of quantum physics, computer science, mathematics, and statistics. Although we do not expect our readers to have the basic knowledge of quantum computing and classical machine learning, we give adequate background and references to accommodate other readers of varying technical background. By 2026, quantum machine learning will be on a important crossroads. NISQ has also provided its own stunning demonstrations and harsh lessons on the difficulty of obtaining useful quantum advantage. Other phenomena like barren plateaus have changed the minds of us on the nature of variational quantum algorithm trainability. Resource-efficient quantum learning has been transformed by the advanced measurement protocols, such as classical shadow tomography. The frontier of the implementable quantum algorithms is constantly being pushed by hardware advances. The book reflects the picture of the field in this dynamic time, including the latest theoretical developments, their experiments, and new optimal practices. We admit that quantum machine learning is a fast-developing science, and there are certain spheres of our knowledge that will surely become more profound in the nearest future. The transformative potential of quantum machine learning is still achievable only with further achievements in many dimensions: the theory of when and why quantum algorithms are effective, algorithmic innovations to overcome the limitations of near-term hardware, hardware engineering to overcome qubit quality and system scale, and discovering applications when quantum algorithms and methods and methods offer significant superior practical performance. Hopefully, this book will provide useful guidance to all the existing knowledge and guide to the future innovations. The opportunities are out of the world, yet the challenges are significant. With the ongoing development of quantum hardware capabilities and an ever-increasing theoretical basis, it is quite possible that quantum machine learning will transform the world of approaching the problem of complex computational learning.
Quantum machine learning (QML) is rapidly transitioning from theoretical promise to practical relevance across data-intensive scientific domains. In this review, we provide a structured overview of recent advances that bridge foundational quantum learning principles with real-world applications. We survey foundational QML paradigms, including variational quantum algorithms, quantum kernel methods, and neural-network quantum states, with emphasis on their applicability to complex quantum systems. We examine neural-network quantum states as expressive variational models for correlated matter, non-equilibrium dynamics, and open quantum systems, and discuss fundamental challenges associated with training and sampling. Recent advances in quantum-enhanced sampling and diagnostics of learning dynamics, including information-theoretic tools, are reviewed as mechanisms for improving scalability and trainability. The review further highlights application-driven QML frameworks in drug discovery, cancer biology, and agro-climate modeling, where data complexity and constraints motivate hybrid quantum–classical approaches. We conclude with a discussion of federated quantum machine learning as a route to distributed, privacy-preserving quantum learning. Overall, this review presents a unified perspective on the opportunities and limitations of QML for complex systems.
This paper delves into the nascent field of quantum machine learning (QML) by establishing a theoretical framework for understanding and predicting the generalization capabilities of QML algorithms. Currently, the practical success of QML algorithms is often attributed to empirical observation, lacking a robust theoretical underpinning. This work addresses this gap by leveraging concepts from quantum information theory to analyze the generalization performance of various QML algorithms. We propose a model that connects algorithm parameters, dataset characteristics, and the resulting generalization error, expressed mathematically as: Γ(θ, D) = ED[|f(x) - ψ(x)|] where: Γ(θ, D) represents the generalization error; θ represents the algorithm's parameters; D represents the dataset; ED[ ] denotes the expected value with respect to the dataset D; f(x) is the model's prediction; ψ(x) is the true label or target. Furthermore, we explore the potential for quantum advantage in generalization, considering the impact of quantum entanglement and superposition on reducing the risk of overfitting. The core contribution lies in providing a methodological approach for quantifying and improving generalization performance in QML, moving beyond purely experimental evaluations. We introduce a novel metric, the Quantum Generalization Index (QGI), defined as: QGI = (1 - Γ(θ, D)) / (1 + Γ(θ, D)) This index provides a normalized measure of generalization performance, allowing for comparative analysis of different QML algorithms and parameter settings. Ultimately, this research aims to provide a foundational understanding for the design and development of more reliable and robust quantum machine learning systems.