Bücher veröffentlicht von Springer US

Reinforcement learning is the learning of a mapping from situations to actions so as to maximize a scalar reward or reinforcement signal. The learner is not told which action to take, as in most forms of machine learning, but instead must discover which actions yield the highest reward by trying them. In the most interesting and challenging cases, actions may affect not only the immediate reward, but also the next situation, and through that all subsequent rewards. These two characteristics -- trial-and-error search and delayed reward -- are the most important distinguishing features of reinforcement learning. Reinforcement learning is both a new and a very old topic in AI. The term appears to have been coined by Minsk (1961), and independently in control theory by Walz and Fu (1965). The earliest machine learning research now viewed as directly relevant was Samuel's (1959) checker player, which used temporal-difference learning to manage delayed reward much as it is used today. Of course learning and reinforcement have been studied in psychology for almost a century, and that work has had a very strong impact on the AI/engineering work. One could in fact consider all of reinforcement learning to be simply the reverse engineering of certain psychological learning processes (e.g. operant conditioning and secondary reinforcement). Reinforcement Learning is an edited volume of original research, comprising seven invited contributions by leading researchers.

The layout of an integrated circuit (lC) is the process of assigning geometric shape, size and position to the components (transistors and connections) used in its fabrication. Since the number of components in modem ICs is enormous, computer- aided-design (CAD) programs are required to automate the difficult layout process. Prior CAD methods are inexact or limited in scope, and produce layouts whose area, and consequently manufacturing costs, are larger than necessary. This book addresses the problem of minimizing exactly the layout area of an important class of basic IC structures called CMOS cells. First, we precisely define the possible goals in area minimization for such cells, namely width and height minimization, with allowance for area-reducing reordering of transistors. We reformulate the layout problem in terms of a graph model and develop new graph-theoretic concepts that completely characterize the fundamental area minimization problems for series-parallel and nonseries-parallel circuits. These concepts lead to practical algorithms that solve all the basic layout minimization problems exactly, both for a single cell and for a one-dimensional array of such cells. Although a few of these layout problems have been solved or partially solved previously, we present here the first complete solutions to all the problems of interest.

Integrated circuit technology is widely used for the full integration of electronic systems. In general, these systems are realized using digital techniques implemented in CMOS technology. The low power dissipation, high packing density, high noise immunity, ease of design and the relative ease of scaling are the driving forces of CMOS technology for digital applications. Parts of these systems cannot be implemented in the digital domain and will remain analog. In order to achieve complete system integration these analog functions are preferably integrated in the same CMOS technology. An important class of analog circuits that need to be integrated in CMOS are analog filters. This book deals with very high frequency (VHF) filters, which are filters with cut-off frequencies ranging from the low megahertz range to several hundreds of megahertz. Until recently the maximal cut-off frequencies of CMOS filters were limited to the low megahertz range. By applying the techniques presented in this book the limit could be pushed into the true VHF domain, and integrated VHF filters become feasible. Application of these VHF filters can be found in the field of communication, instrumentation and control systems. For example, pre and post filtering for high-speed AD and DA converters, signal reconstruction, signal decoding, etc. The general design philosophy used in this book is to allow only the absolute minimum of signal carrying nodes throughout the whole filter. This strategy starts at the filter synthesis level and is extended to the level of electronic circuitry. The result is a filter realization in which all capacitators (including parasitics) have a desired function. The advantage of this technique is that high frequency parasitic effects (parasitic poles/zeros) are minimally present. The book is a reference for engineers in research or development, and is suitable for use as a text for advanced courses on the subject. >

Field-Programmable Gate Arrays (FPGAs) have emerged as an attractive means of implementing logic circuits, providing instant manufacturing turnaround and negligible prototype costs. They hold the promise of replacing much of the VLSI market now held by mask-programmed gate arrays. FPGAs offer an affordable solution for customized VLSI, over a wide variety of applications, and have also opened up new possibilities in designing reconfigurable digital systems. Field-Programmable Gate Arrays discusses the most important aspects of FPGAs in a textbook manner. It provides the reader with a focused view of the key issues, using a consistent notation and style of presentation. It provides detailed descriptions of commercially available FPGAs and an in-depth treatment of the FPGA architecture and CAD issues that are the subjects of current research. The material presented is of interest to a variety of readers, including those who are not familiar with FPGA technology, but wish to be introduced to it, as well as those who already have an understanding of FPGAs, but who are interested in learning about the research directions that are of current interest.

It is a great honor to provide an introduction for Dr. Frank Op 't Eynde's and Dr. Willy Sansen's book "e;Analog Interfaces for Digital Signal Processing Systems"e;. The field of analog integrated circuit design is undergoing rapid evolution. The pervasiveness of digital processing has considerably modified the micro-system architectures: the analog part of complex mixed systems is more and more pushed at the boundary limits of the processing chain. Moreover, the increased performance of digital circuits, in terms of accuracy and speed, are making the specification requirements of analog circuits very strict. In addition to this, the technology, supply voltage and power consumption of analog circuits must be compatible with those, typical for digital circuits. Therefore, in a few words, analog circuits are becoming complex and specialised interfaces between the real world and digital signal processing domains. This technological evolution should be accompanied by an equivalently fast evolution in designer competencies. Knowledge of complicated signal handling should be quickly replaced by know-how of simple but very accurate and very fast signal processing and a solid background in data conversion techniques. All of this through the use of the CMOS (and possibly BiCMOS) technology.

Satellite Communications: Mobile and Fixed Services is based on the premise that designers of future satellite systems must take account of the strong competition that satellites face from optical fibers. In future years, satellites will continue to be commercially viable media for telecommunications only if systems designers take account of the unique features that satellites have to offer. Accordingly, Satellite Communications places more emphasis on satellite mobile services and broadcasting, and less emphasis on fixed, point-to-point, high-capacity services than traditional textbooks in the field. Also, an emphasis is given in the book to design issues. Numerous illustative system design examples and numerical problems are provided. The particular attention given to methods of design of satellite mobile communications systems should make it an indispensable resource for workers in this field. The book also contains some recent results of propagation modelling and system design studies which should be of particular value to researchers and designers of satellite systems for mobile communications services. Satellite Communications is suitable for use as a textbook for advanced courses on satellite communications, and is a valuable reference for all those working in the field.

Generating Abstraction Hierarchies presents a completely automated approach to generating abstractions for problem solving. The abstractions are generated using a tractable, domain-independent algorithm whose only inputs are the definition of a problem space and the problem to be solved and whose output is an abstraction hierarchy that is tailored to the particular problem. The algorithm generates abstraction hierarchies that satisfy the `ordered monotonicity' property, which guarantees that the structure of an abstract solution is not changed in the process of refining it. An abstraction hierarchy with this property allows a problem to be decomposed such that the solution in an abstract space can be held invariant while the remaining parts of a problem are solved. The algorithm for generating abstractions is implemented in a system called ALPINE, which generates abstractions for a hierarchical version of the PRODIGY problem solver. Generating Abstraction Hierarchies formally defines this hierarchical problem solving method, shows that under certain assumptions this method can reduce the size of a search space from exponential to linear in the solution size, and describes the implementation of this method in PRODIGY. The abstractions generated by ALPINE are tested in multiple domains on large problem sets and are shown to produce shorter solutions with significantly less search than problem solving without using abstraction. Generating Abstraction Hierarchies will be of interest to researchers in machine learning, planning and problem reformation.

As MOS devices are scaled to meet increasingly demanding circuit specifications, process variations have a greater effect on the reliability of circuit performance. For this reason, statistical techniques are required to design integrated circuits with maximum yield. Statistical Modeling for Computer-Aided Design of MOS VLSI Circuits describes a statistical circuit simulation and optimization environment for VLSI circuit designers. The first step toward accomplishing statistical circuit design and optimization is the development of an accurate CAD tool capable of performing statistical simulation. This tool must be based on a statistical model which comprehends the effect of device and circuit characteristics, such as device size, bias, and circuit layout, which are under the control of the circuit designer on the variability of circuit performance. The distinctive feature of the CAD tool described in this book is its ability to accurately model and simulate the effect in both intra- and inter-die process variability on analog/digital circuits, accounting for the effects of the aforementioned device and circuit characteristics. Statistical Modeling for Computer-Aided Design of MOS VLSI Circuits serves as an excellent reference for those working in the field, and may be used as the text for an advanced course on the subject.

As any history student will tell you, all events must be understood within their political and sociological context. Yet science provides an interesting counterpoint to this idea, since scientific ideas stand on their own merit, and require no reference to the time and place of their conception beyond perhaps a simple citation. Even so, the historical context of a scientific discovery casts a special light on that discovery - a light that motivates the work and explains its significance against a backdrop of related ideas. The book that you hold in your hands is unusually adept at presenting technical ideas in the context of their time. On one level, Larry Bookman has produced a manuscript to satisfy the requirements of a PhD program. If that was all he did, my preface would praise the originality of his ideas and attempt to summarize their significance. But this book is much more than an accomplished disser- tation about some aspect of natural language - it is also a skillfully crafted tour through a vast body of computational, linguistic, neurophysiological, and psychological research.