The number of models available in neural network literature is quite large. Very often the treatment is mathematical and complex. This book provides illustrative examples in C++ that the reader can use as a basis for further experimentation. A key to learning about neural networks to appreciate their inner workings is to experiment. Neural networks, in the end, are fun to learn about and discover. Although the language for description used is C++, you will not find extensive class libraries in this book. With the exception of the backpropagation simulator, you will find fairly simple example programs for many different neural network architectures and paradigms. Since backpropagation is widely used and also easy to tame, a simulator is provided with the capacity to handle large input data sets. You use the simulator in one of the chapters in this book to solve a financial forecasting problem. You will find ample room to expand and experiment with the code presented in this book.

Computers, Chess, and Cognition presents an excellent up-to-date description of developments in computer chess, a rapidly advancing area in artificial intelligence research. This book is intended for an upper undergraduate and above level audience in the computer science (artificial intelligence) community. The chapters have been edited to present a uniform terminology and balanced writing style, to make the material understandable to a wider, less specialized audience. The book's primary strengths are the description of the workings of some major chess programs, an excellent review of tree searching methods, discussion of exciting new research ideas, a philosophical discussion of the relationship of computer game playing to artificial intelligence, and the treatment of computer Go as an important new research area. A complete index and extensive bibliography makes the book a valuable reference work. The book includes a special foreword by Ken Thompson, author of the UNIX operating system.

This book contains simple examples to provide illustration. The data used in the examples contains information that is not based on any real companies or people. The author and publisher of this book have used their best efforts in preparing this book. The author and publisher make no warranty of any kind, expressed or implied, with regard to the documentation and examples contained in this book. The author and publisher shall not be liable in any event for incidental or consequential damages in connection with, or arising out of the use of, information in this book. The views expressed in this book are the views of the author and do not necessarily reflect the views of the IBM Corporation. This book is not sponsored in any part, or in any manner, by IBM.

Roger Schank's influential book, Dynamic Memory (CUP 1982), described how computers could learn based on what was known about how people learn. Since that book's publication, Dr. Schank has turned his focus from artificial intelligence to human intelligence. Dynamic Memory Revisited contains the theory of learning presented in the original book, extending it to provide principles for teaching and learning, and includes Dr. Schank's important theory of case-based reasoning and assesses the role of stories in human memory. Dynamic Memory Revisited is crucial reading for all who are concerned with education and school reform. It draws attention to how effective learning takes place and provides instruction for developing software that truly helps students learn.

Knowledge is not absolute. This book is not the last word; it is invitation to discussion of understanding and direction of the new class systems development. It can be used to develop the specification for Artificial Intelligent System design and applications. Collective work and evolution of knowledge will create better understanding of these systems philosophy.

Gaussian Scale-Space Theory in Copenhagen, in spring 1996. After this, Bart ter Haar Romeny arranged the First International Conference on Scale-Space Theory in Computer Vision (Utrecht, July 2-4, 1997). Indeed the title was ap- appropriate since this was the first scale-space conference in a series of so far two conferences. We feel very confident that the series will be much longer. We hope that the scheduling next to ICCV '99 will attract more delegates furthering the integration of scale-space theories into computer vision.

Signal Analysis explores methods that offer an insight into the properties of signals and stochastic processes. This comprehensive and authoritative book provides an introduction to the fundamental theory of signal analysis. It bridges the gap between the basic knowledge in system theory and the specialist knowledge in various fields of signal processing and provides a complete overview of current applications. Features include: Focus on both classical and modern approaches to transforms, such as the Karhunen - Lohve transform and the discrete Fourier transform Enhancement of the understanding and relation of transform theory Discussion of recent and emerging topics, including filter banks with perfect reconstruction time-frequency and wavelets. With its great accuracy and technical merit, Signal Analysis is essential reading for signal processing engineers, researchers, undergraduates, postgraduates and lecturers of electrical and computer engineering.

Suitable for even non-specialists, Signal Integrity - Simplified offers a comprehensive, easy-to-follow look at how physical interconnects affect electrical performance. World-class engineer Eric Bogatin expertly reviews the root causes of the four families of signal integrity problems and offers solutions to design them out early in the design cycle.

Methods of recursive identification deal with the problem of building mathematical models of signals and systems on-line, at the same time as data is being collected. Such methods, which are also known as adaptive algorithms or sequential parameter estimation methods, may be applied to a wide spectrum of online adaptive systems, such as devices for signal processing, prediction, or control and are useful for modeling systems in general. For example, they can be used to analyze the demand for power on an electric generating grid and help the grid adjust to continually changing power needs, or applied to the changing conditions of a papermaking plant, or to monitoring pollution in a river. This book provides a comprehensive and systematic framework for developing, describing, and analyzing such recursive algorithms. It has been carefully designed and organized to meet the needs of readers with different objectives. With a myriad of algorithms now in use, it provides a simple and coherent frame of reference for understanding the subject and will serve as a guide to the large number of choices made available by the advent of inexpensive, powerful digital processors.

Methods of recursive identification deal with the problem of building mathematical models of signals and systems on-line, at the same time as data is being collected. Such methods, which are also known as adaptive algorithms or sequential parameter estimation methods, may be applied to a wide spectrum of online adaptive systems, such as devices for signal processing, prediction, or control and are useful for modeling systems in general. For example, they can be used to analyze the demand for power on an electric generating grid and help the grid adjust to continually changing power needs, or applied to the changing conditions of a papermaking plant, or to monitoring pollution in a river. This book provides a comprehensive and systematic framework for developing, describing, and analyzing such recursive algorithms. It has been carefully designed and organized to meet the needs of readers with different objectives. With a myriad of algorithms now in use, it provides a simple and coherent frame of reference for understanding the subject and will serve as a guide to the large number of choices made available by the advent of inexpensive, powerful digital processors. Readers primarily interested in theory will find a detailed development of convergence analysis and asymptotic distribution results. For graduate students it is a basic introduction to the subject. And for engineers interested in practical applications, the book's earlier theory-oriented chapters are equipped with "user's summaries" that provide direct access to the discussion of practical aspects developed in the final three chapters on implementation and applications.

Fundamentally, computers just deal with numbers. They store letters and other characters by assigning a number for each one. There are hundreds of different encoding systems for mapping characters to numbers, but Unicode promises a single mapping. Unicode enables a single software product or website to be targeted across multiple platforms, languages and countries without re-engineering. It's no wonder that industry giants like Apple, Hewlett-Packard, IBM andMicrosoft have all adopted Unicode.

This book is about filter banks and wavelets. Those are new ways to see and represent signal. They are alternatives to the Fourier transform, using short wavelets insted of long wavelets. We will explain the advantages (and disadvantages!) of the new methods. The final decision will depend on the application itself, the actual signal, and its bandwith. To design and understand wavellets we still use Fourier techniques - the connection between time and frequency. This idea remains at the center of signal processing. Wavelets are "alternatives" rather than replacemets. The classical transforms will survive very well. But the ideas have come quickly forward, to be understood and applied.

  Текст у форматі PDF 11.74 Мб

Книга посвящена актуальному научному направлению -проектированию и исследованию систем типа "искусственный разум". Анализируется структура разумного поведения, дается сравнительный анализ используемых алгоритмов обучения и приемов настройки сложных моделей. Рассматриваются алгоритмы и блок-схемы, описывающие вычислительные программы моделей. Рассчитана на кибернетиков и психологов, интересующихся вопросами построения сложных систем.

  Текст у форматі PDF 1.18 Мб

Монография посвящена рассмотрению математического аппарата количественного и качественного анализа АСУ. Конечные автоматы благодаря их простой реализуемости на ЭВМ имеют значительные преимущества по сравнению с другими моделями. Авторы знакомят читателей с основными достижениями в этой области. Книга рассчитана на разработчиков АСУ и цифровых средств вычислительной техники, на математиков, работающих в области системного математического обеспечения и построения проблемно-ориентированных алгоритмических языков, а также на специалистов по системному анализу и моделированию сложных объектов.

Книга американского ученого освещает основные вопросы общей теории линейных кодов, исследования циклических (двоичных и недвоичных) кодов для метрик Хэмминга и Ли, вычисление параметров оптимальных кодов, а также вопросы построения кодирующих и декодирующих устройств. Теоретические исследования сопровождаются большим числом примеров и задач, что делает книгу интересной и доступной не только для математиков, но и для широкого круга специалистов, связанных с разработкой систем передачи цифровой информации, а также для аспирантов и студентов соответствующих специальностей.