This paper presents a new hybrid optimization strategy for training feedforward neural networks. The algorithm combines gradient-based optimization of nonlinear weights with singular value decomposition (SVD) computation of linear weights in one integrated routine. It is described for the multilayer perceptron (MLP) and radial basis function (RBF) networks and then extended to the local model network (LMN), a new feedforward structure in which a global nonlinear model is constructed from a set of locally valid submodels. Simulation results are presented demonstrating the superiority of the new hybrid training scheme compared to second-order gradient methods. It is particularly effective for the LMN architecture where the linear to nonlinear parameter ratio is large.
A software environment is described which provides facilities at a variety of levels for “animating” algorithms: exposing properties of programs by displaying multiple dynamic views of the program and associated data structures. The system is operational on a network of graphics-based, personal workstations and has been used successfully in several applications for teaching and research in computer science and mathematics. In this paper, we outline the conceptual framework that we have developed for animating algorithms, describe the system that we have implemented, and give several examples drawn from the host of algorithms that we have animated.
A software environment is described which provides facilities at a variety of levels for ~animating" algorithms: exposing properties of programs by displaying multiple dynamic views of the program and associated data structures. The system is operational on a network of graphics-based, personal workstations and has been used successfully in several applications for teaching and research in computer science and mathematics. In this paper, we outline the conceptual framework that we have developed for animating algorithms, describe the system that we have implemented, and give several examples drawn from the host of algorithms that we have animated.
Systems Research CenterDEC's business and technology objectives require a strong research program. The Systems Research Center (SRC) and three other research laboratories are committed to filling that need.SRC began recruiting its first research scientists in l984-their charter, to advance the state of knowledge in all aspects of computer systems research. Our current work includes exploring high-performance personal computing, distributed computing, programming environments, system modelling techniques, specification technology, and tightly-coupled multiprocessors.Our approach to both hardware and software research is to create and use real systems so that we can investigate their properties fully. Complex systems cannot be evaluated solely in the abstract. Based on this belief, our strategy is to demonstrate the technical and practical feasibility of our ideas by building prototypes and using them as daily tools. The experience we gain is useful in the short term in enabling us to refine our designs, and invaluable in the long term in helping us to advance the state of knowledge about those systems. Most of the major advances in information systems have come through this strategy, including time-sharing, the ArpaNet, and distributed personal computing.SRC also performs work of a more mathematical flavor which complements our systems research. Some of this work is in established fields of theoretical computer science, such as the analysis of algorithms, computational geometry, and logics of programming. The rest of this work explores new ground motivated by problems that arise in our systems research.DEC has a strong commitment to communicating the results and experience gained through pursuing these activities. The Company values the improved understanding that comes with exposing and testing our ideas within the research community. SRC will therefore report results in conferences, in professional journals, and in our research report series. We will seek users for our prototype systems among those with whom we have common research interests, and we will encourage collaboration with university researchers. This work may not be copied or reproduced in whole or in part for any commercial purpose. Permission to copy in whole or in part without payment of fee is granted for nonprofit educational and research purposes provided that all such whole or partial copies include the following: a notice that such copying is by permission of the Systems Research Center of Digital Equipment Corporation in Palo Alto, California; an acknowledgment of the authors and individual contributors to the work; and all applicable portions of the copyright notice. Copying, reproducing, or republishing for any other purpose shall require a license with payment of fee to the Systems Research Center. All rights reserved.
AbstractAlgorithm animation is a form of program visualization that is concerned with dynamic and interactive graphical displays of a program's fundamental operations. This paper describes the Zeus algorithm animation syste...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.