Machine Learning (ML) is the process of training machines to analyze and learn from provided data. It can be categorized into three main types: supervised, unsupervised, and reinforcement learning. Unsupervised learning, a method that allows the discovery of underlying patterns in data without the need for additional information or labeled targets, uncovers hidden patterns within the given dataset but lacks associated labels or classes. Unsupervised techniques prove particularly useful for exploring data and comprehending complex behaviors that challenge human identification within large datasets. They find applications in various fields, such as text categorization (e.g., news articles), anomaly detection, satellite and spatial image processing, medical image analysis, customer segmentation, and recommendation engines. Unsupervised learning primarily serves three main tasks: clustering, association, and dimensionality reduction. Clustering is a technique that automatically groups similar samples together based on their inherent characteristics. Association is a technique used to discover relationships between different features within a dataset. Dimensionality reduction is a technique employed to decrease the number of features in a dataset, especially when dealing with high-dimensional data.
Machine Learning (ML) is the process of training machines to analyze and learn from provided data. It can be categorized into three main types: supervised, unsupervised, and reinforcement learning. Unsupervised learning, a method that allows the discovery of underlying patterns in data without the need for additional information or labeled targets, uncovers hidden patterns within the given dataset but lacks associated labels or classes. Unsupervised techniques prove particularly useful for exploring data and comprehending complex behaviors that challenge human identification within large datasets. They find applications in various fields, such as text categorization (e.g., news articles), anomaly detection, satellite and spatial image processing, medical image analysis, customer segmentation, and recommendation engines. Unsupervised learning primarily serves three main tasks: clustering, association, and dimensionality reduction. Clustering is a technique that automatically groups similar samples together based on their inherent characteristics. Association is a technique used to discover relationships between different features within a dataset. Dimensionality reduction is a technique employed to decrease the number of features in a dataset, especially when dealing with high-dimensional data.
Quantum machine learning (QML) represents a cutting-edge convergence of quantum computing and artificial intelligence, aiming to leverage quantum phenomena such as superposition, entanglement, and quantum parallelism to enhance machine learning algorithms. By exploiting the computational advantages of quantum systems, QML offers the potential for exponential speedups in data processing, optimization, and pattern recognition tasks, particularly for high-dimensional and complex datasets. This article explores the foundational principles, algorithmic approaches, and practical applications of quantum machine learning across domains including finance, chemistry, drug discovery, optimization, and natural language processing. It discusses quantum-enhanced supervised and unsupervised learning, quantum neural networks, variational quantum algorithms, and hybrid quantum-classical architectures. Additionally, the article addresses challenges related to hardware limitations, noise, scalability, and interpretability, as well as the emerging trends in near-term quantum devices, quantum simulators, and integration with classical AI pipelines. Quantum machine learning holds the promise of revolutionizing computational intelligence, offering transformative capabilities for solving problems that are currently intractable for classical systems.
[EN] In a world in which accessible information grows exponentially, the selection of the appropriate information turns out to be an extremely relevant problem. In this context, the idea of Machine Learning (ML), a subfield of Artificial Intelligence, emerged to face problems in data mining, pattern recognition, automatic prediction, among others. Quantum Machine Learning is an interdisciplinary research area combining quantum mechanics with methods of ML, in which quantum properties allow for an exponential speed-up in the algorithms. This field is currently one of the main research areas in companies like Google and Microsoft, due to the revolution that they could provide in data management.
Bias triangles represent features in stability diagrams of Quantum Dot (QD) devices, whose occurrence and property analysis are crucial indicators for spin physics. Nevertheless, challenges associated with quality and availability of data as well as the subtlety of physical phenomena of interest have hindered an automatic and bespoke analysis framework, often still relying (in part) on human labelling and verification. We introduce a feature extraction framework for bias triangles, built from unsupervised, segmentation-based computer vision methods, which facilitates the direct identification and quantification of physical properties of the former. Thereby, the need for human input or large training datasets to inform supervised learning approaches is circumvented, while additionally enabling the automation of pixelwise shape and feature labeling. In particular, we demonstrate that Pauli Spin Blockade (PSB) detection can be conducted effectively, efficiently and without any training data as a direct result of this approach.
Humans are experts at recognising patterns in past experience and applying them to new tasks. \nFor example, after seeing pictures of a face we can usually tell if another image contains the \nsame person or not. Machine learning is a research discipline at the intersection of computer \nscience, statistics and mathematics that investigates how pattern recognition can be performed \nby machines and for large amounts of data. Since a few years machine learning has come \ninto the focus of quantum computing in which information processing based on the laws of \nquantum theory is explored. Although large scale quantum computers are still in the first stages \nof development, their theoretical description is well-understood and can be used to formulate \n`quantum software' or `quantum algorithms' for pattern recognition. Researchers can therefore \nanalyse the impact quantum computers may have on intelligent data mining. This approach is \npart of the emerging research discipline of quantum machine learning that harvests synergies \nbetween quantum computing and machine learning. \nThe research objective of this thesis is to understand how we can solve a slightly more specific \nproblem called supervised pattern recognition based on the language that has been developed \nfor universal quantum computers. The contribution it makes is twofold: First, it presents a \nmethodology that understands quantum machine learning as the combination of data encoding into \nquantum systems and quantum optimisation. Second, it proposes several quantum algorithms for \nsupervised pattern recognition. These include algorithms for convex and non-convex optimisation, \nimplementations of distance-based methods through quantum interference, and the preparation of \nquantum states from which solutions can be derived via sampling. Amongst the machine learning \nmethods considered are least-squares linear regression, gradient descent and Newton's method, \nk-nearest neighbour, neural networks as well as ensemble methods. Together with the growing \nbody of literature, this thesis demonstrates that quantum computing offers a number of interesting \ntools for machine learning applications, and has the potential to create new models of how to learn \nfrom data.
Quantum Machine Learning investigates the pos-sibility of quantum computers enhancing Machine Learning algorithms. Anomaly segmentation is a fundamental task in various domains to identify irregularities at sample level and can be addressed with both supervised and unsupervised methods. Autoencoders are commonly used in unsupervised tasks, where models are trained to reconstruct normal instances efficiently, al-lowing anomaly identification through high reconstruction errors. While quantum autoencoders have been proposed in the literature, their application to anomaly segmentation tasks remains unexplored. In this paper, we introduce a patch-based quantum autoencoder (QPB-AE) for image anomaly segmentation, with a number of parameters scaling logarithmically with patch size. QPB-AE reconstructs the quantum state of the embedded input patches, computing an anomaly map directly from measurement through a SWAP test without reconstructing the input image. We evaluate its performance across multiple datasets and parameter configurations and compare it against a classical counterpart.
In medical applications, machine learning often grapples with limited training data. Classical self-supervised deep learning techniques have been helpful in this domain, but these algorithms have yet to achieve the required accuracy for medical use. Recently quantum algorithms show promise in handling complex patterns with small datasets. To address this challenge, this study presents a novel solution that combines self-supervised learning with Variational Quantum Classifiers (VQC) and utilizes Principal Component Analysis (PCA) as the dimensionality reduction technique. This unique approach ensures generalization even with a small training dataset while preserving data privacy, a vital consideration in medical applications. PCA is effectively utilized for dimensionality reduction, enabling VQC to operate with just 2 Q-bits, overcoming current quantum hardware limitations, and gaining an advantage over classical methods. The proposed model was benchmarked against linear classification models using diverse public image datasets to validate its effectiveness. The results demonstrate remarkable accuracy, with achievements of 90% on PneumoniaMNIST, 90% on BreastMNIST, 80% on PathMNIST, and 80% on ChestMNIST medical datasets. Additionally, for non-medical datasets, the model attained 85% on Hymenoptera Ants & Bees and 90% on the Kaggle Cats & Dogs dataset.