This book provides a comprehensive set of characterization, prediction, optimization, evaluation, and evolution techniques for a diagnosis system for fault isolation in large electronic systems. Readers with a background in electronics design or system engineering can use this book as a reference to derive insightful knowledge from data analysis and use this knowledge as guidance for designing reasoning-based diagnosis systems. Moreover, readers with a background in statistics or data analytics can use this book as a practical case study for adapting data mining and machine learning techniques to electronic system design and diagnosis. This book identifies the key challenges in reasoning-based, board-level diagnosis system design and presents the solutions and corresponding results that have emerged from leading-edge research in this domain. It covers topics ranging from highly accurate fault isolation, adaptive fault isolation, diagnosis-system robustness assessment, to system performance analysis and evaluation, knowledge discovery and knowledge transfer. With its emphasis on the above topics, the book provides an in-depth and broad view of reasoning-based fault diagnosis system design. * Explains and applies optimized techniques from the machine-learning domain to solve the fault diagnosis problem in the realm of electronic system design and manufacturing;* Demonstrates techniques based on industrial data and feedback from an actual manufacturing line;* Discusses practical problems, including diagnosis accuracy, diagnosis time cost, evaluation of diagnosis system, handling of missing syndromes in diagnosis, and need for fast diagnosis-system development.

This book tackles important problems of anomaly detection and health status analysis in complex core router systems, integral to today's Internet Protocol (IP) networks. The techniques described provide the first comprehensive set of data-driven resiliency solutions for core router systems. The authors present an anomaly detector for core router systems using correlation-based time series analysis, which monitors a set of features of a complex core router system. They also describe the design of a changepoint-based anomaly detector such that anomaly detection can be adaptive to changes in the statistical features of data streams. The presentation also includes a symbol-based health status analyzer that first encodes, as a symbol sequence, the long-term complex time series collected from a number of core routers, and then utilizes the symbol sequence for health analysis. Finally, the authors describe an iterative, self-learning procedure for assessing the health status. Enables Accurate Anomaly Detection Using Correlation-Based Time-Series Analysis; Presents the design of a changepoint-based anomaly detector; Includes Hierarchical Symbol-based Health-Status Analysis; Describes an iterative, self-learning procedure for assessing the health status.

This book provides a comprehensive set of characterization, prediction, optimization, evaluation, and evolution techniques for a diagnosis system for fault isolation in large electronic systems. Readers with a background in electronics design or system engineering can use this book as a reference to derive insightful knowledge from data analysis and use this knowledge as guidance for designing reasoning-based diagnosis systems. Moreover, readers with a background in statistics or data analytics can use this book as a practical case study for adapting data mining and machine learning techniques to electronic system design and diagnosis. This book identifies the key challenges in reasoning-based, board-level diagnosis system design and presents the solutions and corresponding results that have emerged from leading-edge research in this domain. It covers topics ranging from highly accurate fault isolation, adaptive fault isolation, diagnosis-system robustness assessment, to system performance analysis and evaluation, knowledge discovery and knowledge transfer. With its emphasis on the above topics, the book provides an in-depth and broad view of reasoning-based fault diagnosis system design. * Explains and applies optimized techniques from the machine-learning domain to solve the fault diagnosis problem in the realm of electronic system design and manufacturing;* Demonstrates techniques based on industrial data and feedback from an actual manufacturing line;* Discusses practical problems, including diagnosis accuracy, diagnosis time cost, evaluation of diagnosis system, handling of missing syndromes in diagnosis, and need for fast diagnosis-system development.

This book tackles important problems of anomaly detection and health status analysis in complex core router systems, integral to today's Internet Protocol (IP) networks. The techniques described provide the first comprehensive set of data-driven resiliency solutions for core router systems. The authors present an anomaly detector for core router systems using correlation-based time series analysis, which monitors a set of features of a complex core router system. They also describe the design of a changepoint-based anomaly detector such that anomaly detection can be adaptive to changes in the statistical features of data streams. The presentation also includes a symbol-based health status analyzer that first encodes, as a symbol sequence, the long-term complex time series collected from a number of core routers, and then utilizes the symbol sequence for health analysis. Finally, the authors describe an iterative, self-learning procedure for assessing the health status. Enables Accurate Anomaly Detection Using Correlation-Based Time-Series Analysis; Presents the design of a changepoint-based anomaly detector; Includes Hierarchical Symbol-based Health-Status Analysis; Describes an iterative, self-learning procedure for assessing the health status.