Doctoral Thesis / Dissertation from the year 2010 in the subject Computer Science - Applied, grade: none, , course: Department of Computers and Information Sciences - Ph.D., language: English, abstract: Although Support Vector Machines have been used to develop highly accurate classification and regression models in various real-world problem domains, the most significant barrier is that SVM generates black box model that is difficult to understand. The procedure to convert these opaque models into transparent models is called rule extraction. This thesis investigates the task of extracting comprehensible models from trained SVMs, thereby alleviating this limitation. The primary contribution of the thesis is the proposal of various algorithms to overcome the significant limitations of SVM by taking a novel approach to the task of extracting comprehensible models. The basic contribution of the thesis are systematic review of literature on rule extraction from SVM, identifying gaps in the literature and proposing novel approaches for addressing the gaps. The contributions are grouped under three classes, decompositional, pedagogical and eclectic/hybrid approaches. Decompositional approach is closely intertwined with the internal workings of the SVM. Pedagogical approach uses SVM as an oracle to re-label training examples as well as artificially generated examples. In the eclectic/hybrid approach, a combination of these two methods is adopted. The thesis addresses various problems from the finance domain such as bankruptcy prediction in banks/firms, churn prediction in analytical CRM and Insurance fraud detection. Apart from this various benchmark datasets such as iris, wine and WBC for classification problems and auto MPG, body fat, Boston housing, forest fires and pollution for regression problems are also tested using the proposed appraoch. In addition, rule extraction from unbalanced datasets as well as from active learning based approaches has been explored. For classification problems, various rule extraction methods such as FRBS, DT, ANFIS, CART and NBTree have been utilized. Additionally for regression problems, rule extraction methods such as ANFIS, DENFIS and CART have also been employed. Results are analyzed using accuracy, sensitivity, specificity, fidelity, AUC and t-test measures. Proposed approaches demonstrate their viability in extracting accurate, effective and comprehensible rule sets in various benchmark and real world problem domains across classification and regression problems. Future directions have been indicated to extend the approaches to newer variations of SVM as well as to other problem domains.
Presenting comprehensive and up-to-date state of the domain of highly strained hydrocarbons with unusual spatial structure, an experienced editor and top authors cover the whole range of these important molecules from [1.1.1]propellane to fullerenes and nanotubes. The necessity of studies in this area encompassing sometimes exotic molecules is discussed in detail showing their importance for basic science and practical applications. The fact that the latter cannot mostly be foreseen is amply documented. Even not long ago, studying such molecules was an elitist activity, few synthetic chemists succeeded in their syntheses. Today, the field has broadened in view of emerging practical application and fullerenes and nanotubes are one of the most vividly developing domains. Chapters include both experimental and theoretical studies. The former cover syntheses and unusual physicochemical properties related to the strain and untypical geometry of the molecules under scrutiny. The latter show importance of model calculations which help precise basic ideas of chemistry, such as the chemical bond, on the one hand, and are used to propose novel plausible synthetic targets, on the other. The monograph aims not only at PhD students and newcomers in the field who seek for an introduction to this area but also at the specialists who want to obtain a broader perspective of this domain and make use of a comprehensive review of literature.
A billiard is a dynamical system in which a point particle alternates between free motion and specular reflections from the boundary of a domain. Exterior Billiards presents billiards in the complement of domains and their applications in aerodynamics and geometrical optics. This book distinguishes itself from existing literature by presenting billiard dynamics outside bounded domains, including scattering, resistance, invisibility and retro-reflection. It begins with an overview of the mathematical notations used throughout the book and a brief review of the main results. Chapters 2 and 3 are focused on problems of minimal resistance and Newton&#8217;s problem in media with positive temperature. In chapters 4 and 5, scattering of billiards by nonconvex and rough domains is characterized and some related special problems of optimal mass transportation are studied. Applications in aerodynamics are addressed next and problems of invisibility and retro-reflection within the framework of geometric optics conclude the text. The book will appeal to mathematicians working in dynamical systems and calculus of variations. Specialists working in the areas of applications discussed will also find it useful.
Provides principles and best practices for the design and development of enterprise software applications Enterprise software drives much of the world's IT systems in critical domains such as healthcare, finance, e-commerce, and government. Web services and cloud computing are two manifestations of the growing sophistication and interconnectedness of business IT processes. Despite its importance, the agreed best practices for building enterprise software are still very much a 'work in progress.' This text provides a timely review of the principles and practice of enterprise software architecture, including the perspectives of software-oriented architecture (SOA), domain-driven design, and representational state transfer (REST). It also features: * Both the implementation-oriented perspective of the hands-on developer and the design-oriented perspective of the software architect * Discussion of the support for enterprise software development provided by popular frameworks such as Java Enterprise Edition and Windows Communication Foundation, including illustrative examples * Discussion of trends in computer science research that promise to have a bearing on the hard problems associated with building reliable enterprise applications * Review of basic concepts that any enterprise software developer or architect will need to understand, in programming languages and distributed systems Intended to bridge the gap between high-level conceptual overviews and in-depth technical tutorials, Enterprise Software Architecture and Design is ideal for students in computer science, information systems, and systems engineering, as well as software development professionals, computer scientists, and software architects.
'This is currently the best book covering the relationship between genome and computer architectures.' - JOHNATHAN BARTLETT, Author / Publisher / Speaker / Director of Technology ----- This book highlights the informational aspects of life that are generally overlooked or ignored in chemical and biological evolutionary scenarios. Each cell of an organism has millions of interacting computers reading and processing digital information, using digital programs and digital codes to communicate and translate information. Life is an intersection of physical science and information science. Both domains are critical for any life to exist, and each must be investigated using that domain's principles. Yet most scientists have been attempting to use physical science to explain life's information domain, a practice which has no scientific justification. -- As you can tell by the preceding words this research is a fascinating approach to the question of the origin of life. - (PUBLISHER) ----- 'Programming of Life is an excellent freshman level review of the formal programming, coding/decoding, integration, organization, Prescriptive Information (PI), memory, regulation and control required for a physical object to find itself 'alive.' DONALD E. JOHNSON is uniquely qualified to unpackage the strong parallels between everyday cybernetic design and engineering and the workings of the cell. I highly recommend this book.' -DAVID L. ABEL, Director, The Gene Emergence Project Department of ProtoBioCybernetics and ProtoBioSemiotics The Origin of Life Science Foundation, Inc. ----- (ABOUT THE AUTHOR:) DR. DON JOHNSON has earned Ph.D.s in both Computer & Information Sciences from the University of Minnesota and in Chemistry from Michigan State University. He was a senior research scientist for 10 years in pharmaceutical and medical / scientific instrument fields, served as president and technical expert in an independent computer consulting firm for many years, and taught for 20 years in universities in Wisconsin, Minnesota, California, and Europe. He now maintains scienceintegrity.net to expose unsubstantiated claims in science and has made presentations on most continents.
Kernel methods have long been established as effective techniques in the framework of machine learning and pattern recognition, and have now become the standard approach to many remote sensing applications. With algorithms that combine statistics and geometry, kernel methods have proven successful across many different domains related to the analysis of images of the Earth acquired from airborne and satellite sensors, including natural resource control, detection and monitoring of anthropic infrastructures (e.g. urban areas), agriculture inventorying, disaster prevention and damage assessment, and anomaly and target detection. Presenting the theoretical foundations of kernel methods (KMs) relevant to the remote sensing domain, this book serves as a practical guide to the design and implementation of these methods. Five distinct parts present state-of-the-art research related to remote sensing based on the recent advances in kernel methods, analysing the related methodological and practical challenges: * Part I introduces the key concepts of machine learning for remote sensing, and the theoretical and practical foundations of kernel methods. * Part II explores supervised image classification including Super Vector Machines (SVMs), kernel discriminant analysis, multi-temporal image classification, target detection with kernels, and Support Vector Data Description (SVDD) algorithms for anomaly detection. * Part III looks at semi-supervised classification with transductive SVM approaches for hyperspectral image classification and kernel mean data classification. * Part IV examines regression and model inversion, including the concept of a kernel unmixing algorithm for hyperspectral imagery, the theory and methods for quantitative remote sensing inverse problems with kernel-based equations, kernel-based BRDF (Bidirectional Reflectance Distribution Function), and temperature retrieval KMs. * Part V deals with kernel-based feature extraction and provides a review of the principles of several multivariate analysis methods and their kernel extensions. This book is aimed at engineers, scientists and researchers involved in remote sensing data processing, and also those working within machine learning and pattern recognition.
Boost your Linux+/LPIC readiness with practice tests for all exam domains CompTIA Linux+ and LPIC Practice Tests provide 100% coverage of all exam objectives for both the CompTIA Linux+ exams LX0-103 and LX0-104, and the LPIC exams 101-400, 102-400, and 201and 202, through 1,200+ expertly crafted practice questions. These easy to navigate practice questions cover the Linux+ and LPIC-1 exams, covering all 10 domains. The second part covers the LPIC-2 exam, covering all 13 LPIC-2 domains. An additional two 60-question practice exams per section help you gauge your readiness -- and hone your test-taking strategy -- well in advance of exam day, giving you the thorough preparation you need to approach the exam with confidence. You will also gain access to the Sybex interactive learning environment where you have access to all questions and can create your own practice tests based on areas further review is needed Master the skills and concepts on the LPIC-1 and the LPIC-2 exams Gauge your understanding with unique practice tests for each exam domain Identify weak points and prioritize your preparation Preview exam day with four 60-question practice exams Practice tests are among the most effective exam preparation strategies. These tests are designed to mimic the Linux+, LPIC-1 and LPIC-2 exams, giving you the practice you need to ensure you are fully prepared. This book can be used alone or with the Sybex study guides and deluxe study guides. Start preparing for your Linux certification today.
Wireless sensor/actuators networks (WSANs) are being increasingly used in a panoply of applications, such as industrial automation, process control, ambient assisted living, structural health monitoring, and homeland security. Most of these applications require specific quality-of-service (QoS) guarantees from their underlying communication infrastructures (regardless of their wireless, wired, or hybrid nature). This book gathers together an extremely rich set of contributions, addressing several WSAN domains and sharing QoS as a common denominator. Eight papers have made it through a rigorous and iterative peer review process (three reviews per paper, at least two review rounds), involving 38 authors from all over the world (North and South America, Europe, Asia, and Australia) from academia, industry, and the military. Each paper features at least one reference author which is highly reputed in this scientific domain, totaling over 100,000 citations altogether.
Data registration refers to a series of techniques for matching or bringing similar objects or datasets together into alignment. These techniques enjoy widespread use in a diverse variety of applications, such as video coding, tracking, object and face detection and recognition, surveillance and satellite imaging, medical image analysis and structure from motion. Registration methods are as numerous as their manifold uses, from pixel level and block or feature based methods to Fourier domain methods. This book is focused on providing algorithms and image and video techniques for registration and quality performance metrics. The authors provide various assessment metrics for measuring registration quality alongside analyses of registration techniques, introducing and explaining both familiar and state-of-the-art registration methodologies used in a variety of targeted applications. Key features: * Provides a state-of-the-art review of image and video registration techniques, allowing readers to develop an understanding of how well the techniques perform by using specific quality assessment criteria * Addresses a range of applications from familiar image and video processing domains to satellite and medical imaging among others, enabling readers to discover novel methodologies with utility in their own research * Discusses quality evaluation metrics for each application domain with an interdisciplinary approach from different research perspectives