Machine learning

This book addresses the growing need for machine learning and data mining in neuroscience. The book is replete with fully working machine learning code. It also contains lab assignments and quizzes, making it appropriate for use as a textbook.

The concept of deep machine learning is easier to understand by paying attention to the cyclic stochastic time series and a time series whose content is non-stationary not only within the cycles, but also over the cycles as the cycle-to-cycle variations.

In this book, innovative research using artificial neural networks (ANNs) is conducted to automate the sizing task of RF IC design, which is used in two different steps of the automatic design process. The advances in telecommunications, such as the 5th generation broadband or 5G for short, open doors to advances in areas such as health care, education, resource management, transportation, agriculture and many other areas. Consequently, there is high pressure in today¿s market for significant communication rates, extensive bandwidths and ultralow-power consumption. This is where radiofrequency (RF) integrated circuits (ICs) come in hand, playing a crucial role. This demand stresses out the problem which resides in the remarkable difficulty of RF IC design in deep nanometric integration technologies due to their high complexity and stringent performances. Given the economic pressure for high quality yet cheap electronics and challenging time-to-market constraints, there is an urgent need for electronic design automation (EDA) tools to increase the RF designers¿ productivity and improve the quality of resulting ICs. In the last years, the automatic sizing of RF IC blocks in deep nanometer technologies has moved toward process, voltage and temperature (PVT)-inclusive optimizations to ensure their robustness. Each sizing solution is exhaustively simulated in a set of PVT corners, thus pushing modern workstations¿ capabilities to their limits.Standard ANNs applications usually exploit the model¿s capability of describing a complex, harder to describe, relation between input and target data. For that purpose, ANNs are a mechanism to bypass the process of describing the complex underlying relations between data by feeding it a significant number of previously acquired input/output data pairs that the model attempts to copy. Here, and firstly, the ANNs disrupt from the most recent trials of replacing the simulator in the simulation-based sizing with a machine/deep learning model, by proposing two different ANNs, the first classifies the convergence of the circuit for nominal and PVT corners, and the second predicts the oscillating frequencies for each case. The convergence classifier (CCANN) and frequency guess predictor (FGPANN) are seamlessly integrated into the simulation-based sizing loop, accelerating the overall optimization process. Secondly, a PVT regressor that inputs the circuit¿s sizing and the nominal performances to estimate the PVT corner performances via multiple parallel artificial neural networks is proposed. Two control phases prevent the optimization process from being misled by inaccurate performance estimates. As such, this book details the optimal description of the input/output data relation that should be fulfilled. The developed description is mainly reflected in two of the system¿s characteristics, the shape of the input data and its incorporation in the sizing optimization loop. An optimal description of thesecomponents should be such that the model should produce output data that fulfills the desired relation for the given training data once fully trained. Additionally, the model should be capable of efficiently generalizing the acquired knowledge in newer examples, i.e., never-seen input circuit topologies.

Machine Learning Algorithms in Depth dives deep into the 'how' and the 'why' of machine learning algorithms. For each category of an algorithm, you will go from math-first principles to hands-on implementation in Python. You will explore dozens of examples from across all the fields of machine learning, including finance, computer vision, NLP, and more. Each example is accompanied by worked-out derivations and details as well as insightful code samples and graphics. By the time you're done reading, you will know how major algorithms work under the hood -- and be a better machine learning practitioner. About the reader For intermediate machine learning practitioners familiar with linear algebra, probability, and basic calculus.

This book constitutes the Third 3D Head and Neck Tumor Segmentation in PET/CT Challenge, HECKTOR 2022, which was held in conjunction with the 25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022, on September 22, 2022.The 22 contributions presented, as well as an overview paper, were carefully reviewed and selected from 24 submissions. This challenge aims to evaluate and compare the current state-of-the-art methods for automatic head and neck tumor segmentation. In the context of this challenge, a dataset of 883 delineated PET/CT images was made available for training.

This open access book provides an introduction and an overview of learning to quantify (a.k.a. ¿quantification¿), i.e. the task of training estimators of class proportions in unlabeled data by means of supervised learning. In data science, learning to quantify is a task of its own related to classification yet different from it, since estimating class proportions by simply classifying all data and counting the labels assigned by the classifier is known to often return inaccurate (¿biased¿) class proportion estimates.The book introduces learning to quantify by looking at the supervised learning methods that can be used to perform it, at the evaluation measures and evaluation protocols that should be used for evaluating the quality of the returned predictions, at the numerous fields of human activity in which the use of quantification techniques may provide improved results with respect to the naive use of classification techniques, and at advanced topics in quantification research.The book is suitable to researchers, data scientists, or PhD students, who want to come up to speed with the state of the art in learning to quantify, but also to researchers wishing to apply data science technologies to fields of human activity (e.g., the social sciences, political science, epidemiology, market research) which focus on aggregate (¿macrö) data rather than on individual (¿micrö) data.

This book constitutes three challenges that were held in conjunction with the 25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022, which took place in Singapore in September 2022. The peer-reviewed 10 papers included in this volume stem from the following three challenges: Kidney Parsing Challenge 2022: Multi-Structure Segmentation for Renal Cancer Treatment (KiPA 2022) The 2022 Correction of Brain Shift with Intra-Operative Ultrasound-Segmentation Challenge (CuRIOUS-SEG 2022) The 2022 Mediastinal Lesion Analysis Challenge (MELA 2022)

This book constitutes the refereed post-conference proceedings of the workshops held at the 16th Asian Conference on Computer Vision, ACCV 2022, which took place in Macao, China, in December 2022. The 25 papers included in this book were carefully reviewed and selected from 40 submissions. They have been organized in topical sections as follows: Learning with limited data for face analysis; adversarial machine learning towards advanced vision systems; computer vision for medical computing; machine learning and computing for visual semantic analysis; vision transformers theory and applications; and deep learning-based small object detection from images and videos.

This book introduces recent research results for cyber deception, a promising field for proactive cyber defense. The beauty and challenge of cyber deception is that it is an interdisciplinary research field requiring study from techniques and strategies to human aspects. This book covers a wide variety of cyber deception research, including game theory, artificial intelligence, cognitive science, and deception-related technology. Specifically, this book addresses three core elements regarding cyber deception: Understanding human¿s cognitive behaviors in decoyed network scenarios Developing effective deceptive strategies based on human¿s behaviorsDesigning deceptive techniques that supports the enforcement of deceptive strategiesThe research introduced in this book identifies the scientific challenges, highlights the complexity and inspires the future research of cyber deception.Researchers working in cybersecurity and advanced-level computer science students focused on cybersecurity will find this book useful as a reference. This book also targets professionals working in cybersecurity.Chapter 'Using Amnesia to Detect Credential Database Breaches' and Chapter 'Deceiving ML-Based Friend-or-Foe Identification for Executables' are available open access under a Creative Commons Attribution 4.0 International License via link.springer.com.

This book deals with behavioral responses of management of firms that make several decisions with respect to production, marketing, finance, organization of activities within divisions, and interrelations between divisions (including synergies between them and constraints placed on each other in the attainment of overall goals of the firm). The market conditions, that constitute the basis of such decisions, may be stable, random but predictable, or uncertain. It can be expected that objectives attained by the firm, as a result of decisions of management, may be different from the maximum which can be achieved. A generic conceptualization of such managerial discretion and operationally useful methods of measurement have been presented. It is possible to develop machine learning algorithms on this basis to minimize managerial discretion and assist managers in arriving at strategic decisions thereby leaving more resources to deal with uncertain events as they arise. The volume is a great resource not only for researchers, but also decision makers in corporates.

This book explains how to scale Apache Spark 3 to handle massive amounts of data, either via batch or streaming processing. It covers how to use Spark¿s structured APIs to perform complex data transformations and analyses you can use to implement end-to-end analytics workflows. This book covers Spark 3's new features, theoretical foundations, and application architecture. The first section introduces the Apache Spark ecosystem as a unified engine for large scale data analytics, and shows you how to run and fine-tune your first application in Spark. The second section centers on batch processing suited to end-of-cycle processing, and data ingestion through files and databases. It explains Spark DataFrame API as well as structured and unstructured data with Apache Spark. The last section deals with scalable, high-throughput, fault-tolerant streaming processing workloads to process real-time data. Here you'll learn about Apache Spark Streaming¿s execution model, the architecture of Spark Streaming, monitoring, reporting, and recovering Spark streaming. A full chapter is devoted to future directions for Spark Streaming. With real-world use cases, code snippets, and notebooks hosted on GitHub, this book will give you an understanding of large-scale data analysis concepts--and help you put them to use.Upon completing this book, you will have the knowledge and skills to seamlessly implement large-scale batch and streaming workloads to analyze real-time data streams with Apache Spark.What You Will LearnMaster the concepts of Spark clusters and batch data processingUnderstand data ingestion, transformation, and data storageGain insight into essential stream processing concepts and different streaming architecturesImplement streaming jobs and applications with Spark StreamingWho This Book Is ForData engineers, data analysts, machine learning engineers, Python and R programmers

A converter with a single input and multiple outputs is known as a multi-output converter. The converter could be isolated or non-isolated. A non-isolated converter which uses a single common inductor for all the outputs is termed single-inductor multi-output converter. In a multi-output converter, a part of the converter circuit is common for all the outputs and a part of the circuit is exclusive to each output. This brings in cost advantages and improved power density. Fewer components also result in reduced losses. Multi-output converters have become very popular recently and many low power and portable applications widely use such converters. Since a part of the converter is common for all the outputs, achange in the load of one output affects the other outputs also. This is known as cross-regulation and the converters have to handle this apart from line and load regulation. This is what makes the topic complex and interesting and provides scope for further investigation, despite the large volume of research work already existing in this area. This thesis attempts to identify new topologies and control approaches for configuring such multi-output converters in all spheres, isolated or non-isolated, converters operating in discontinuous conduction mode or in continuous conduction mode and with uni-polar or bi-polar outputs.

This book presents the modern technological advancements and revolutions in the biomedical sector. Progress in the contemporary sensing, Internet of Things (IoT) and machine learning algorithms and architectures have introduced new approaches in the mobile healthcare. A continuous observation of patients with critical health situation is required. It allows monitoring of their health status during daily life activities such as during sports, walking and sleeping. It is realizable by intelligently hybridizing the modern IoT framework, wireless biomedical implants and cloud computing. Such solutions are currently under development and in testing phases by healthcare and governmental institutions, research laboratories and biomedical companies. The biomedical signals such as electrocardiogram (ECG), electroencephalogram (EEG), Electromyography (EMG), phonocardiogram (PCG), Chronic Obstructive Pulmonary (COP), Electrooculography (EoG), photoplethysmography (PPG), and image modalitiessuch as positron emission tomography (PET), magnetic resonance imaging (MRI) and computerized tomography (CT) are non-invasively acquired, measured, and processed via the biomedical sensors and gadgets. These signals and images represent the activities and conditions of human cardiovascular, neural, vision and cerebral systems. Multi-channel sensing of these signals and images with an appropriate granularity is required for an effective monitoring and diagnosis. It renders a big volume of data and its analysis is not feasible manually. Therefore, automated healthcare systems are in the process of evolution. These systems are mainly based on biomedical signal and image acquisition and sensing, preconditioning, features extraction and classification stages. The contemporary biomedical signal sensing, preconditioning, features extraction and intelligent machine and deep learning-based classification algorithms are described. Each chapter starts with the importance, problemstatement and motivation. A self-sufficient description is provided. Therefore, each chapter can be read independently. To the best of the editors¿ knowledge, this book is a comprehensive compilation on advances in non-invasive biomedical signal sensing and processing with machine and deep learning. We believe that theories, algorithms, realizations, applications, approaches, and challenges, which are presented in this book will have their impact and contribution in the design and development of modern and effective healthcare systems.

This book includes the proceedings of the fourth workshop on recommender systems in fashion and retail (2022), and it aims to present a state-of-the-art view of the advancements within the field of recommendation systems with focused application to e-commerce, retail, and fashion by presenting readers with chapters covering contributions from academic as well as industrial researchers active within this emerging new field. Recommender systems are often used to solve different complex problems in this scenario, such as product recommendations, size and fit recommendations, and social media-influenced recommendations (outfits worn by influencers).

This book presents the state-of-the-art applications of machine learning in the finance domain with a focus on financial product modeling, which aims to advance the model performance and minimize risk and uncertainty. It provides both practical and managerial implications of financial and managerial decision support systems which capture a broad range of financial data traits. It also serves as a guide for the implementation of risk-adjusted financial product pricing systems, while adding a significant supplement to the financial literacy of the investigated study.The book covers advanced machine learning techniques, such as Support Vector Machine, Neural Networks, Random Forest, K-Nearest Neighbors, Extreme Learning Machine, Deep Learning Approaches, and their application to finance datasets. It also leverages real-world financial instances to practice business product modeling and data analysis. Software code, such as MATLAB, Python and/or R including datasets within a broad range of financial domain are included for more rigorous practice.The book primarily aims at providing graduate students and researchers with a roadmap for financial data analysis. It is also intended for a broad audience, including academics, professional financial analysts, and policy-makers who are involved in forecasting, modeling, trading, risk management, economics, credit risk, and portfolio management.

This book presents intelligent methods like neural, neuro-fuzzy, machine learning, deep learning and metaheuristic methods and their applications in both volcanology and seismology. The complex system of volcanoes and also earthquakes is a big challenge to identify their behavior using available models, which motivates scientists to apply non-model based methods. As there are lots of seismology and volcanology data sets, i.e., the local and global networks, one solution is using intelligent methods in which data-based algorithms are used.

The definitive guide to the game-theoretic and probabilistic underpinning for Bitcoin's security model. The book begins with an overview of probability and game theory. Nakamoto Consensus is discussed in both practical and theoretical terms. This volume: Describes attacks and exploits with mathematical justifications, including selfish mining. Identifies common assumptions such as the Market Fragility Hypothesis, establishing a framework for analyzing incentives to attack. Outlines the block reward schedule and economics of ASIC mining. Discusses how adoption by institutions would fundamentally change the security model. Analyzes incentives for double-spend and sabotage attacks via stock-flow models. Overviews coalitional game theory with applications to majority takeover attacks Presents Nash bargaining with application to unregulated environments This book is intended for students or researchers wanting to engage in a serious conversation about the future viability of Bitcoin as a decentralized, censorship-resistant, peer-to-peer electronic cash system.

In den letzten Jahren ist die Sportinformatik extrem gewachsen, vor allem weil immer mehr und neuere Daten verfügbar wurden. Sportinformatische Tools ¿ sei es im Training zur Gegnervorbereitung, im Wettkampf oder in der Wissenschaft ¿ sind im Sport heute auf unterschiedlichen Expertise-Ebenen unverzichtbar. Durch den Einsatz in den vier großen Anwendungsfeldern Vereine und Verbände, Wirtschaft, Wissenschaft sowie Medien ist ein völlig neuer Markt entstanden, der innerhalb der universitären Forschungs- und Lehraktivitäten zunehmend an Bedeutung gewinnt. Dieses Lehrbuch möchte der mittlerweile breiten Vielfalt der Sportinformatik gerecht werden, indem mehr als 30 Autorinnen und Autoren aus ihrem Spezialgebiet berichten und neueste Erkenntnisse prägnant zusammenfassen. Das Werk gliedert sich in vier Hauptabschnitte: Datensätze, Modellbildung, Simulation sowie Datenanalyse. Neben Hintergründen zu Programmiersprachen und zur Visualisierung wird es von der Historie und einem Ausblick eingerahmt. Studierende mit Bezug zur Sportwissenschaft erhalten einen umfassenden Einblick in die Sportinformatik, unterstützt durch ein didaktisch ausgefeiltes Konzept, das eine einfache Vermittlung der Lerninhalte ermöglicht. Zahlreiche digitale Übungsfragen untermauern den Lerneffekt und gewährleisten eine optimale Prüfungsvorbereitung. Für Fortgeschrittene bietet die vertiefende Diskussion von Zeitreihen Data Mining, künstlichen neuronalen Netzwerken, Convolution Kernel, Transfer Learning und Random Forests einen zusätzlichen Mehrwert.

This book includes selected papers presented at the International Conference on Marketing and Technologies (ICMarkTech 2021), held at University of La Laguna, Tenerife, Spain, during December 2-4, 2021. It covers up-to-date cutting-edge research on artificial intelligence applied in marketing, virtual and augmented reality in marketing, business intelligence databases and marketing, data mining and big data, marketing data science, web marketing, e-commerce and v-commerce, social media and networking, geomarketing and IoT, marketing automation and inbound marketing, machine learning applied to marketing, customer data management and CRM, and neuromarketing technologies.