Machine Learning under Resource Constraints addresses novel machine learning algorithms that are challenged by high-throughput data, by high dimensions, or by complex structures of the data in three volumes. Resource constraints are given by the relation between the demands for processing the data and the capacity of the computing machinery. The resources are runtime, memory, communication, and energy. Hence, modern computer architectures play a significant role. Novel machine learning algorithms are optimized with regard to minimal resource consumption. Moreover, learned predictions are executed on diverse architectures to save resources. It provides a comprehensive overview of the novel approaches to machine learning research that consider resource constraints, as well as the application of the described methods in various domains of science and engineering. Volume 1 establishes the foundations of this new field. It goes through all the steps from data collection, their summary and clustering, to the different aspects of resource-aware learning, i.e., hardware, memory, energy, and communication awareness. Several machine learning methods are inspected with respect to their resource requirements and how to enhance their scalability on diverse computing architectures ranging from embedded systems to large computing clusters.
Machine learning is part of Artificial Intelligence since its beginning. Certainly, not learning would only allow the perfect being to show intelligent behavior. All others, be it humans or machines, need to learn in order to enhance their capabilities. In the eighties of the last century, learning from examples and modeling human learning strategies have been investigated in concert. The formal statistical basis of many learning methods has been put forward later on and is still an integral part of machine learning. Neural networks have always been in the toolbox of methods. Integrating all the pre-processing, exploitation of kernel functions, and transformation steps of a machine learning process into the architecture of a deep neural network increased the performance of this model type considerably. Modern machine learning is challenged on the one hand by the amount of data and on the other hand by the demand of real-time inference. This leads to an interest in computing architectures and modern processors. For a long time, the machine learning research could take the von-Neumann architecture for granted. All algorithms were designed for the classical CPU. Issues of implementation on a particular architecture have been ignored. This is no longer possible. The time for independently investigating machine learning and computational architecture is over. Computing architecture has experienced a similarly rampant development from mainframe or personal computers in the last century to now very large compute clusters on the one hand and ubiquitous computing of embedded systems in the Internet of Things on the other hand. Cyber-physical systems’ sensors produce a huge amount of streaming data which need to be stored and analyzed. Their actuators need to react in real-time. This clearly establishes a close connection with machine learning. Cyber-physical systems and systems in the Internet of Things consist of diverse components, heterogeneous both in hard- and software. Modern multi-core systems, graphic processors, memory technologies and hardware-software codesign offer opportunities for better implementations of machine learning models. Machine learning and embedded systems together now form a field of research which tackles leading edge problems in machine learning, algorithm engineering, and embedded systems. Machine learning today needs to make the resource demands of learning and inference meet the resource constraints of used computer architecture and platforms. A large variety of algorithms for the same learning method and, moreover, diverse implementations of an algorithm for particular computing architectures optimize learning with respect to resource efficiency while keeping some guarantees of accuracy. The trade-off between a decreased energy consumption and an increased error rate, to just give an example, needs to be theoretically shown for training a model and the model inference. Pruning and quantization are ways of reducing the resource requirements by either compressing or approximating the model. In addition to memory and energy consumption, timeliness is an important issue, since many embedded systems are integrated into large products that interact with the physical world. If the results are delivered too late, they may have become useless. As a result, real-time guarantees are needed for such systems. To efficiently utilize the available resources, e.g., processing power, memory, and accelerators, with respect to response time, energy consumption, and power dissipation, different scheduling algorithms and resource management strategies need to be developed. This book series addresses machine learning under resource constraints as well as the application of the described methods in various domains of science and engineering. Turning big data into smart data requires many steps of data analysis: methods for extracting and selecting features, filtering and cleaning the data, joining heterogeneous sources, aggregating the data, and learning predictions need to scale up. The algorithms are challenged on the one hand by high-throughput data, gigantic data sets like in astrophysics, on the other hand by high dimensions like in genetic data. Resource constraints are given by the relation between the demands for processing the data and the capacity of the computing machinery. The resources are runtime, memory, communication, and energy. Novel machine learning algorithms are optimized with regard to minimal resource consumption. Moreover, learned predictions are applied to program executions in order to save resources. The three books will have the following subtopics: Volume 1: Machine Learning under Resource Constraints - Fundamentals Volume 2: Machine Learning and Physics under Resource Constraints - Discovery Volume 3: Machine Learning under Resource Constraints - Applications Volume 3 describes how the resource-aware machine learning methods and techniques are used to successfully solve real-world problems. The book provides numerous specific application examples. In the areas of health and medicine, it is demonstrated how machine learning can improve risk modelling, diagnosis, and treatment selection for diseases. Machine learning supported quality control during the manufacturing process in a factory allows to reduce material and energy cost and save testing times is shown by the diverse real-time applications in electronics and steel production as well as milling. Additional application examples show, how machine-learning can make traffic, logistics and smart cities more efficient and sustainable. Finally, mobile communications can benefit substantially from machine learning, for example by uncovering hidden characteristics of the wireless channel.
Machine learning is part of Artificial Intelligence since its beginning. Certainly, not learning would only allow the perfect being to show intelligent behavior. All others, be it humans or machines, need to learn in order to enhance their capabilities. In the eighties of the last century, learning from examples and modeling human learning strategies have been investigated in concert. The formal statistical basis of many learning methods has been put forward later on and is still an integral part of machine learning. Neural networks have always been in the toolbox of methods. Integrating all the pre-processing, exploitation of kernel functions, and transformation steps of a machine learning process into the architecture of a deep neural network increased the performance of this model type considerably. Modern machine learning is challenged on the one hand by the amount of data and on the other hand by the demand of real-time inference. This leads to an interest in computing architectures and modern processors. For a long time, the machine learning research could take the von-Neumann architecture for granted. All algorithms were designed for the classical CPU. Issues of implementation on a particular architecture have been ignored. This is no longer possible. The time for independently investigating machine learning and computational architecture is over. Computing architecture has experienced a similarly rampant development from mainframe or personal computers in the last century to now very large compute clusters on the one hand and ubiquitous computing of embedded systems in the Internet of Things on the other hand. Cyber-physical systems’ sensors produce a huge amount of streaming data which need to be stored and analyzed. Their actuators need to react in real-time. This clearly establishes a close connection with machine learning. Cyber-physical systems and systems in the Internet of Things consist of diverse components, heterogeneous both in hard- and software. Modern multi-core systems, graphic processors, memory technologies and hardware-software codesign offer opportunities for better implementations of machine learning models. Machine learning and embedded systems together now form a field of research which tackles leading edge problems in machine learning, algorithm engineering, and embedded systems. Machine learning today needs to make the resource demands of learning and inference meet the resource constraints of used computer architecture and platforms. A large variety of algorithms for the same learning method and, moreover, diverse implementations of an algorithm for particular computing architectures optimize learning with respect to resource efficiency while keeping some guarantees of accuracy. The trade-off between a decreased energy consumption and an increased error rate, to just give an example, needs to be theoretically shown for training a model and the model inference. Pruning and quantization are ways of reducing the resource requirements by either compressing or approximating the model. In addition to memory and energy consumption, timeliness is an important issue, since many embedded systems are integrated into large products that interact with the physical world. If the results are delivered too late, they may have become useless. As a result, real-time guarantees are needed for such systems. To efficiently utilize the available resources, e.g., processing power, memory, and accelerators, with respect to response time, energy consumption, and power dissipation, different scheduling algorithms and resource management strategies need to be developed. This book series addresses machine learning under resource constraints as well as the application of the described methods in various domains of science and engineering. Turning big data into smart data requires many steps of data analysis: methods for extracting and selecting features, filtering and cleaning the data, joining heterogeneous sources, aggregating the data, and learning predictions need to scale up. The algorithms are challenged on the one hand by high-throughput data, gigantic data sets like in astrophysics, on the other hand by high dimensions like in genetic data. Resource constraints are given by the relation between the demands for processing the data and the capacity of the computing machinery. The resources are runtime, memory, communication, and energy. Novel machine learning algorithms are optimized with regard to minimal resource consumption. Moreover, learned predictions are applied to program executions in order to save resources. The three books will have the following subtopics: Volume 1: Machine Learning under Resource Constraints - Fundamentals Volume 2: Machine Learning and Physics under Resource Constraints - Discovery Volume 3: Machine Learning under Resource Constraints - Applications Volume 2 is about machine learning for knowledge discovery in particle and astroparticle physics. Their instruments, e.g., particle accelerators or telescopes, gather petabytes of data. Here, machine learning is necessary not only to process the vast amounts of data and to detect the relevant examples efficiently, but also as part of the knowledge discovery process itself. The physical knowledge is encoded in simulations that are used to train the machine learning models. At the same time, the interpretation of the learned models serves to expand the physical knowledge. This results in a cycle of theory enhancement supported by machine learning.
A new edition of a graduate-level machine learning textbook that focuses on the analysis and theory of algorithms. This book is a general introduction to machine learning that can serve as a textbook for graduate students and a reference for researchers. It covers fundamental modern topics in machine learning while providing the theoretical basis and conceptual tools needed for the discussion and justification of algorithms. It also describes several key aspects of the application of these algorithms. The authors aim to present novel theoretical tools and concepts while giving concise proofs even for relatively advanced topics. Foundations of Machine Learning is unique in its focus on the analysis and theory of algorithms. The first four chapters lay the theoretical foundation for what follows; subsequent chapters are mostly self-contained. Topics covered include the Probably Approximately Correct (PAC) learning framework; generalization bounds based on Rademacher complexity and VC-dimension; Support Vector Machines (SVMs); kernel methods; boosting; on-line learning; multi-class classification; ranking; regression; algorithmic stability; dimensionality reduction; learning automata and languages; and reinforcement learning. Each chapter ends with a set of exercises. Appendixes provide additional material including concise probability review. This second edition offers three new chapters, on model selection, maximum entropy models, and conditional entropy models. New material in the appendixes includes a major section on Fenchel duality, expanded coverage of concentration inequalities, and an entirely new entry on information theory. More than half of the exercises are new to this edition.
Machine Learning: A Constraint-Based Approach provides readers with a refreshing look at the basic models and algorithms of machine learning, with an emphasis on current topics of interest that includes neural networks and kernel machines. The book presents the information in a truly unified manner that is based on the notion of learning from environmental constraints. While regarding symbolic knowledge bases as a collection of constraints, the book draws a path towards a deep integration with machine learning that relies on the idea of adopting multivalued logic formalisms, like in fuzzy systems. A special attention is reserved to deep learning, which nicely fits the constrained- based approach followed in this book. This book presents a simpler unified notion of regularization, which is strictly connected with the parsimony principle, and includes many solved exercises that are classified according to the Donald Knuth ranking of difficulty, which essentially consists of a mix of warm-up exercises that lead to deeper research problems. A software simulator is also included.
Master the essentials of project management with this guide to delivering results on time and within budget. Covering key methodologies and best practices, this book is an invaluable resource for project managers aiming for successful project outcomes.
Deep learning networks are getting smaller. Much smaller. The Google Assistant team can detect words with a model just 14 kilobytes in size—small enough to run on a microcontroller. With this practical book you’ll enter the field of TinyML, where deep learning and embedded systems combine to make astounding things possible with tiny devices. Pete Warden and Daniel Situnayake explain how you can train models small enough to fit into any environment. Ideal for software and hardware developers who want to build embedded systems using machine learning, this guide walks you through creating a series of TinyML projects, step-by-step. No machine learning or microcontroller experience is necessary. Build a speech recognizer, a camera that detects people, and a magic wand that responds to gestures Work with Arduino and ultra-low-power microcontrollers Learn the essentials of ML and how to train your own models Train models to understand audio, image, and accelerometer data Explore TensorFlow Lite for Microcontrollers, Google’s toolkit for TinyML Debug applications and provide safeguards for privacy and security Optimize latency, energy usage, and model and binary size
The fundamental mathematical tools needed to understand machine learning include linear algebra, analytic geometry, matrix decompositions, vector calculus, optimization, probability and statistics. These topics are traditionally taught in disparate courses, making it hard for data science or computer science students, or professionals, to efficiently learn the mathematics. This self-contained textbook bridges the gap between mathematical and machine learning texts, introducing the mathematical concepts with a minimum of prerequisites. It uses these concepts to derive four central machine learning methods: linear regression, principal component analysis, Gaussian mixture models and support vector machines. For students and others with a mathematical background, these derivations provide a starting point to machine learning texts. For those learning the mathematics for the first time, the methods help build intuition and practical experience with applying mathematical concepts. Every chapter includes worked examples and exercises to test understanding. Programming tutorials are offered on the book's web site.
Methods by which robots can learn control laws that enable real-time reactivity using dynamical systems; with applications and exercises. This book presents a wealth of machine learning techniques to make the control of robots more flexible and safe when interacting with humans. It introduces a set of control laws that enable reactivity using dynamical systems, a widely used method for solving motion-planning problems in robotics. These control approaches can replan in milliseconds to adapt to new environmental constraints and offer safe and compliant control of forces in contact. The techniques offer theoretical advantages, including convergence to a goal, non-penetration of obstacles, and passivity. The coverage of learning begins with low-level control parameters and progresses to higher-level competencies composed of combinations of skills. Learning for Adaptive and Reactive Robot Control is designed for graduate-level courses in robotics, with chapters that proceed from fundamentals to more advanced content. Techniques covered include learning from demonstration, optimization, and reinforcement learning, and using dynamical systems in learning control laws, trajectory planning, and methods for compliant and force control . Features for teaching in each chapter: applications, which range from arm manipulators to whole-body control of humanoid robots; pencil-and-paper and programming exercises; lecture videos, slides, and MATLAB code examples available on the author’s website . an eTextbook platform website offering protected material[EPS2] for instructors including solutions.
