Machine-learning models often replace vectors by normalized directions, projectors, covariances, subspaces, ordered flags, quantum states, or density operators before any classifier is fitted. This replacement is an invariance decision: it determines which distinctions are kept and which are quotiented out. We develop a self-contained real-to-Hermitian taxonomy for auditing such representations in quantum machine learning. On the real side, we formalize Grassmann and flag projector kernels, prove positive semidefiniteness and block-gauge invariance of a weighted flag kernel, and give a same-span block-swap witness showing when whole-span Grassmann geometry must fail while ordered flags succeed. On the quantum side, we prove that a noiseless fidelity kernel is exactly the Hilbert--Schmidt inner product between the associated rank-one Hermitian projectors, and that a QVR-style return probability is exactly an overlap score between the input projector and a learned anchor operator. Rank-constrained returns correspond to complex Grassmann anchors, while mixed or multimodal class models are naturally represented by density or positive-semidefinite anchors. Controlled vector, subspace, statevector, anomaly, finite-shot, and quotient-witness experiments support the same conclusion: quantum and geometric lifts are useful when their invariances match the task, and fail correctly when discarded information is label-bearing. The paper makes no hardware-speedup or quantum-advantage claim.
Machine-learning models often replace vectors by normalized directions, projectors, covariances, subspaces, ordered flags, quantum states, or density operators before any classifier is fitted. This replacement is an invariance decision: it determines which distinctions are kept and which are quotiented out. We develop a self-contained real-to-Hermitian taxonomy for auditing such representations in quantum machine learning. On the real side, we formalize Grassmann and flag projector kernels, prove positive semidefiniteness and block-gauge invariance of a weighted flag kernel, and give a same-span block-swap witness showing when whole-span Grassmann geometry must fail while ordered flags succeed. On the quantum side, we prove that a noiseless fidelity kernel is exactly the Hilbert--Schmidt inner product between the associated rank-one Hermitian projectors, and that a QVR-style return probability is exactly an overlap score between the input projector and a learned anchor operator. Rank-constrained returns correspond to complex Grassmann anchors, while mixed or multimodal class models are naturally represented by density or positive-semidefinite anchors. Controlled vector, subspace, statevector, anomaly, finite-shot, and quotient-witness experiments support the same conclusion: quantum and geometric lifts are useful when their invariances match the task, and fail correctly when discarded information is label-bearing. The paper makes no hardware-speedup or quantum-advantage claim.
This chapter develops quantum kernel-based machine learning as one of the most mature and practically viable paradigms in quantum machine learning (QML). It explains how kernel methods separate quantum representation from classical learning, allowing quantum devices to define feature spaces through state preparation and kernel estimation while leaving optimization and inference to established classical algorithms. This hybrid structure makes quantum kernel methods especially well suited to near-term quantum hardware and provides a conceptually clean framework for studying the role of quantum-induced geometry in learning. This chapter examines the historical emergence of kernel methods in QML, their positioning relative to variational models, and the practical workflow of learning from precomputed quantum kernels. It then analyzes the issues of generalization, regularization, and kernel noise, showing how finite-shot estimation and hardware imperfections affect the learned geometry of the feature space. Two major models are developed in detail: quantum support vector machines (QSVMs), which perform margin maximization in quantum-induced Hilbert spaces, and quantum kernel ridge regression (QKRR), which provides a closed-form framework for studying stability and conditioning. Together, these models illustrate how quantum kernels can be integrated into standard classical learning pipelines while preserving both mathematical rigor and experimental feasibility.
The quantum support vector machines (QSVMs) offer a paradigm change that may leverage quantum mechanical processes to outperform traditional support vector machines (SVMs) in high-dimensional and nonlinear classification tasks. These machines are located at the crossroads of supervised machine learning and quantum computing. The purpose of this chapter is to provide a comprehensive technical analysis of QSVMs, with a particular emphasis on the use of variation quantum circuits for implicit feature space transformations and the development of these circuits within the context of quantum kernel estimation. Through the use of quantum state preparation and interference, we provide a comprehensive explanation of the possible fabrication of quantum kernels that are based on inner product evaluations in exponentially huge Hilbert spaces. When it comes to boosting model expressivity beyond classical limits, the importance that entanglement and quantum feature mapping play is specifically stressed. In addition to this, we investigate algorithmic implementations on noisy intermediate-scale quantum (NISQ) hardware, focusing on problems associated with quantum decoherence, gate fidelity, and loop depth optimization. Several fields, including quantum-enhanced drug discovery, anomaly detection in quantum sensor networks, fault classification in quantum error correction, and portfolio optimization, are now the focus of research that is being conducted in order to develop real-world applications. This chapter comes to a close with an analysis of the existing theoretical limits, resource needs, and future perspectives for the use of QSVMs in machine learning applications in order to achieve a quantum advantage. In order to facilitate the development of scalable quantum machine learning models that are founded on strong mathematical and computational notions, this synthesis is geared at academics and practitioners.
This chapter develops the theory and practice of quantum feature maps and quantum kernels as a central framework in quantum machine learning (QML). Building on the data-encoding methods introduced in the previous chapter, it explains how classical inputs are embedded into quantum states and how similarity between data points is then defined through overlaps in Hilbert space. This chapter introduces the quantum kernel as the squared overlap between encoded states, establishes its fundamental algebraic properties, and shows how it satisfies the conditions required for direct use in classical kernel-learning methods such as support vector machines, kernel ridge regression, and Gaussian process models. A major focus is placed on the circuit-level realization and estimation of quantum kernels. This chapter presents kernel evaluation through encoding unitaries and examines three principal estimation methods: the SWAP test, the Hadamard test, and the direct return-probability method, with attention to their trade-offs in qubit overhead, circuit depth, and suitability for NISQ hardware. It also develops the geometric interpretation of quantum kernels in Hilbert space, relating kernel values to angular separation, decision boundaries, expressivity, and generalization. Together, these ideas provide a unified foundation for understanding how quantum circuits, feature maps, and classical learning algorithms interact in hybrid quantum-classical kernel models.
Quantum kernel methods retain the standard supervised-learning interface while defining similarities through overlaps of quantum feature states. In practice, the resulting Gram matrices can develop strong global components that dominate the spectrum and make geometric comparisons to classical kernels difficult. This motivates separating global effects from task-relevant variation and testing whether classical postprocessing can reveal and reshape the remaining structure without changing the embedding circuit. Kernel centering (double-centering) is therefore used to remove the constant component while preserving positive semidefiniteness. In parallel, a controlled set of classical extensions is examined at three levels, namely elementwise transforms on datapoints before embedding, spectrum-level eigenvalue mappings, and entrywise kernel transformations, with trace normalization used for consistent comparison. Overall, the results show that the centering reshapes kernel spectra and geometry-based comparisons, often amplifying the separation from a classical RBF reference, even though model predictions are invariant. In contrast, circuit-independent classical extensions can affect predictive performance, narrowing the classification gap for fidelity kernels and improving regression performance, in some settings matching or exceeding the classical baseline. These findings position kernel postprocessing as a practical and low-cost way to analyze quantum kernel geometry and to improve kernel-based supervised learning pipelines (e.g., SVC/SVR) that use quantum kernels.
We study whether quantum learning advantage can persist without fully embedding a large classical input into a highly superposed quantum state. To address this question, we introduce active quantum subspace data-encoding, in which only an information-bearing subset of the input is lifted to a quantum representation while the remaining variables stay classical. For this model, we define a projected hybrid readout and prove three structural results. First, the projected hybrid kernel is positive semidefinite and its sample regularized dimension is bounded by the number of projected observables, so the dimension blow-up of naive global kernels is avoided. Second, we give a necessary and sufficient criterion for improvement over a purely classical predictor in squared loss: the projected quantum sector must contain a direction that lies outside the classical feature span and correlates with the classical residual. Third, in a realizable noisy-oracle setting, we derive a PAC sample-complexity bound proportional to the inverse square of the oracle reliability. We then show, for a canonical Clifford active-subspace family under local dephasing noise, that this reliability can remain inverse-polynomial even when the encoding gate complexity grows polynomially with system size. Hence, the polynomial encoding cost does not by itself destroy the hybrid learning advantage. A sixty-four-qubit family and a synthetic contextual classification task illustrate how one projected quantum feature can compress a useful high-order interaction into a low-dimensional hybrid model. Our results generalize QRAM-free hybrid learning and provide a scalable route toward NISQ-compatible quantum advantage without full quantum data-encoding.
This chapter examines quantum linear algebra as a foundational framework for quantum machine learning. Because many core learning tasks, including regression, dimensionality reduction, kernel evaluation, recommendation systems, and representation learning, depend on linear algebraic operations, this chapter explores how these operations can be reformulated in the quantum setting. Emphasis is placed on the shift from explicit numerical computation to coherent manipulation of quantum states and measurement-based access to information. This chapter introduces quantum inner product estimation as a primitive for similarity and kernel evaluation, highlighting the swap test and related overlap-estimation methods. It then develops quantum matrix inversion and the HHL algorithm, emphasizing the preparation of solution states rather than explicit solution vectors and discussing the importance of conditioning and regularization. Quantum singular value decomposition is presented through Hermitian embeddings, phase estimation, block encoding, and quantum singular value transformation, providing a unified framework for spectral filtering and low-rank approximation. This chapter concludes with quantum principal component analysis, in which covariance structure is represented by a density matrix and accessed through quantum phase estimation and density matrix exponentiation. Throughout, practical demonstrations in Qiskit and PennyLane illustrate the algorithms, while careful attention is given to the limitations imposed by state preparation, measurement, noise, and classical data loading.