Open Access. Powered by Scholars. Published by Universities.®

Engineering Commons™

Open Access. Powered by Scholars. Published by Universities.®

Washington University in St. Louis

Discipline
Keyword
Publication Year
Publication
Publication Type
File Type

Articles 1681 - 1710 of 1825

Full-Text Articles in Engineering

High-Performance Training Of Feedforward & Simple Recurrent Networks, Barry L. Kalman, Stan C. Kwasny Jan 1994

High-Performance Training Of Feedforward & Simple Recurrent Networks, Barry L. Kalman, Stan C. Kwasny

All Computer Science and Engineering Research

TRAINREC is a system for training feedforward and recurrent neural networks that incorporates several ideas. It uses the conjugate-gradient method which is demonstrably more efficient than traditional backward error propagation. We assume epoch-based training and derive a new error function having several desirable properties absent from the traditional sum-of-squared-error function. We argue for skip (shortcut) connections where appropriate and the preference for a sigmoidal yielding values over the [-1,1] interval. The input feature space is often over-analyzed, but by using singular value decomposition, input patterns can be conditioned for better learning often with a reduced number of input units. Recurrent …

Go to article

Learning And Teaching Of Boolean And Geometric Classes, H. David Mathias Jan 1994

Learning And Teaching Of Boolean And Geometric Classes, H. David Mathias

All Computer Science and Engineering Research

We consider the concept classes of DNF formulas and unions of discretized, axis-parallel d-dimensional boxes in discretized d-dimensional space with respect to several different learning models. In the model of learning with queries we present an algorithm to learn unions of boxes. We introduce a model of teaching that prevents illicit communication between the teacher and the leaner but that captures the intuitive aspect of teaching: a learner should perform at least as well with a cooperative teacher as with an adversarial teacher. We propose the study of teaching of DNF formulas and unions of boxes in this model. We …

Go to article

Trading Packet Headers For Packet Processing, George Varghese Jan 1994

Trading Packet Headers For Packet Processing, George Varghese

All Computer Science and Engineering Research

In high speed networks, packet processing is relatively expensive while bandwidth is cheap. This begs the question: what fields can be added to packets to make packet processing easier? By exploring this question, we device a number of novel mechanisms to speed up packet processing. With the advent of new standards for hte Data Link, Network, and Transport lyaers, we believe there is an opportunity to apply these techniques to improve the performance of real protocols. First, we suggest adding a data manipulation header to an easily accessible portion of each packet. This header contains pointers to fields (in various …

Go to article

Rationales And Argument Moves, R. P. Loui, Jeff Norman Jan 1994

Rationales And Argument Moves, R. P. Loui, Jeff Norman

All Computer Science and Engineering Research

No abstract provided.

Go to article

Learning From A Consistently Ignorant Teacher, Michael Frazier, Sally Goldman, Nina Mishra, Leonard Pitt Jan 1994

Learning From A Consistently Ignorant Teacher, Michael Frazier, Sally Goldman, Nina Mishra, Leonard Pitt

All Computer Science and Engineering Research

One view of computational learning theory is that of a learner acquiring the knowledge of a teacher. We introduce a formal model of learning capturing the idea that teachers may have gaps in their knowledge. The goal of the learner is still to acquire the knowledge of the teacher, but now the learner must also identify the gaps. This is the notion of learning from a consistently ignorant teacher. We consider the impact of knowledge gaps on learning, for example, monoton DNF and d-dimensional boxes, and show that leraning is still possible. Negatively, we show that knowledge gaps make learning …

Go to article

Cell Tracking Using A Distributed Algorithm For 3d Image Segmentation, Vikas Awasthi, Keith W. Doolittle, Guru Parulkar, James G. Mcnally Jan 1994

Cell Tracking Using A Distributed Algorithm For 3d Image Segmentation, Vikas Awasthi, Keith W. Doolittle, Guru Parulkar, James G. Mcnally

All Computer Science and Engineering Research

We have developed and tested an automated method for simultaneous 3D tracking of numerous, flourescently-tagged cells. The procedure uses multiple thresholding to segment individual cells at a starting timepoint, and then iteratively applies a template-matching algorithm to locate a particular cell's position at subsequent time points. To speed up the method, we have developed a distributed implementation in which template matching is carried out in parallel on several different server machines. The distributed implementation showed a monotonic decrease in response time with increasing number of servers (up to 15 tested), demonstrating that the tracking algorithm is well suited to parallelization, …

Go to article

Design Of A Large Scale Multimedia Server, Milind M. Buddhikot, Guru Parulkar, Jerome R. Cox Jr. Jan 1994

Design Of A Large Scale Multimedia Server, Milind M. Buddhikot, Guru Parulkar, Jerome R. Cox Jr.

All Computer Science and Engineering Research

Large scale multimedia storage servers will be an integral part of the emerging distributed multimedia computing infrastructure. However, given the modest rate of improvements in storage transfer rates, designing servers that meet the demands of multimedia applications is a challenging task that needs significant architectural innovation. Our research project, called Massively-parallel And Real-time Storage (MARS) architecture, is aimed at the design and prototype implementation of a large scale multimedia storage server. It uses some of the well-known techniques in parallel I/O, such as data striping and Redundant Arrays of Inexpensive Disks (RAID) and an innovative ATM based interconnect inside the …

Go to article

Learning One-Dimensional Geometric Patterns Under One-Sided Random Misclassification Noise, Paul W. Goldberg, Sally A. Goldman Jan 1994

Learning One-Dimensional Geometric Patterns Under One-Sided Random Misclassification Noise, Paul W. Goldberg, Sally A. Goldman

All Computer Science and Engineering Research

Developing the ability to recognize a landmark from a visual image of a robot's current location is a fundamental problem in robotics. We consider the problem of PAC-learning the concept class of geometric patterns where the target geometric pattern is a configuration of k points on the real line. Each instance is a configuration of n points on the real line, where it is labeled according to whether or not it visually resembles the target pattern. To capture the notion of visual resemblance we use the Hausdorff metric. Informally, two geometric patterns P and Q resemble each othe runder the …

Go to article

Catching Up With The Networks: Host I/O At Gigabit Rates, Zubin D. Dittia, Jerome R. Cox Jr., Guru M. Parulkar Jan 1994

Catching Up With The Networks: Host I/O At Gigabit Rates, Zubin D. Dittia, Jerome R. Cox Jr., Guru M. Parulkar

All Computer Science and Engineering Research

The last few years have seen network data rates skyrocket from a few Mbps to a Gbps or more. However, a lack of integration of the host-netowrk interface, the operating system, and network protocols has resulted in end-applications seeing only a small fraction of this total bandwidth being available for data transfer. The emergence of demanding applications in the realms of multimedia and virtual reality provides further impetus in the drive to overcome this problem. In this paper, we present the design of a high performance ATM host-network interface for workstations and servers that can support a bidirecitonal sustained data …

Go to article

Boxgraph: A Two-Dimensional Visual Computation Model, Takayuki Dan Kimura, Timothy B. Brown Jan 1994

Boxgraph: A Two-Dimensional Visual Computation Model, Takayuki Dan Kimura, Timothy B. Brown

All Computer Science and Engineering Research

Traditional computation models such as Turing machines, lambda-calculus, Markov's normal algorithms, are not suitable models for visual programming languages because they are all based on one-dimensional text strings and visual programming uses two-dimensional graphic diagrams. We propose a two-dimensional computation model, called Boxgraph, that requires no text. The syntax of the model consists of nested boxes connected by arrows, and the semantics consists of dataflow and the concept of consistency. The expressive power of the model is demonstrated by constructing representations of a binary full adder, the Fibonacci function, and the GCD function. The model, with a small extension to …

Go to article

Visual Presentation Of Software Specifications And Designs, Gruia-Catalin Roman, Delbert Hart, Charles Calkins Jan 1994

Visual Presentation Of Software Specifications And Designs, Gruia-Catalin Roman, Delbert Hart, Charles Calkins

All Computer Science and Engineering Research

Formal methods hold the promise for high dependability in the design of critical software. However, software engineers who employ formal methods need to communicate their design decisions to users, customers, managers, and collegues who may not be in a position to acquire a full understanding of the formal notation being used. Visualizations derived from formal specifications and designs must be able to convey the required information precisely and reliably without the use of formal notation. This paper discusses a design methodology which attempts to integrate a design methodology based upon specification and program refinement with a state-of-the-art approach to rapid …

Go to article

Speculative Computation: Overcoming Communication Delays In Parallel Algorithms, Vasudha Govindan, Mark A. Franklin Jan 1994

Speculative Computation: Overcoming Communication Delays In Parallel Algorithms, Vasudha Govindan, Mark A. Franklin

All Computer Science and Engineering Research

Communication latencies and delays are a major source of performance degradation in parallel computing systems. It is importnat to "mask" these communication delays by overlapping them with useful computation in order to obtain good parallel performance. This paper proposes speculative computation as a technique to mask communication latencies. Speculative computation is discussed in the context of synchronous iterative algorithms. Processors speculate the contents of messages that are not hyet received and perform computation based on the speculated values. When the messages are received, they are compared with the speculated values and, if the error is unacceptable, the resulitng computation is …

Go to article

Universal Continuous Media I/O: Design And Implementation, Charles D. Cranor, Gurudatta M. Parulkar Jan 1994

Universal Continuous Media I/O: Design And Implementation, Charles D. Cranor, Gurudatta M. Parulkar

All Computer Science and Engineering Research

The problem this paper addresses is how to modify an existing operating system's I/O subsystem to support new high-speed networks and high-bandwidth multimedia applications that will play an important role in future computing environments. The proposed I/O subsystem is called universal continuous media I/O (UCM I/O). This paper will cover the preliminary design of UCM I/O, some of the trade-offs and issues that need to be addressed in order to implement UCM I/O, and a summary of work in progress.

Go to article

Proposal For Research Distribution Of Gigabit Network Technology, Jonathan Turner Jan 1994

Proposal For Research Distribution Of Gigabit Network Technology, Jonathan Turner

All Computer Science and Engineering Research

In 1993, APRA funded a major program at Washingotn University to create gigabit networking technology and create a gigabit testbed based on this technology. This program is now nearing the end of its first year and is making excellent proress towards its research and technical objectives. This note a proposes a program that would lead to the export of this technology to research groups in networking and gigabit applications with an interest in using it to further their own research activities.

Go to article

Connection Management In Reconfigurable Distributed Systems, Bala Swaminathan Jan 1994

Connection Management In Reconfigurable Distributed Systems, Bala Swaminathan

All Computer Science and Engineering Research

The Programmer's Playground takes a new approach to simplifying and supporting the construction of distributed applications. The approach, called I/O abstraction, separates the description of a system's communication structure from the descriptions of its computational components so that software modules written in existing programming languages cna be integrated flexibly and dynamically by both programmers and end-users. This separation is achieved by estabishing logical connectinos among the data interfaces of independent software modules. The logical connections provide a uniform high-level view of communication for both discrete and continuous data. The I/O abstraction approach inherits ideas from the I/O automaton model, a …

Go to article

Efficient Fair Queueing Using Deficit Round Robin, George Varghese, M. Shreedhar Jan 1994

Efficient Fair Queueing Using Deficit Round Robin, George Varghese, M. Shreedhar

All Computer Science and Engineering Research

Fair queuing is a technique that allows each flow passing through a network device to have fair share of network resources. previous schemes for fair queuing that achieved nearly perfect fairness were expensive to implement: specifically, the work required to process a packet in these schemes was O(log(n)), where n is the number of active flows. This is expensive at high speeds. On the other hand, cheaper approximations of fair queuing that have been reported in the literature exhibit unfair behavior. In this paper, we describe a new approximation of fair queuing, that we call Deficit Round Robin. Our scheme …

Go to article

Cmap, Ken Cox, John Dehart Jan 1994

Cmap, Ken Cox, John Dehart

All Computer Science and Engineering Research

This document specifies a Connection Management Access Protocol (CMAP) for call management in high-speed packet switched networks. We target CMAP to networks employing the Asynchronous Transfer Mode (ATM) communication standard. CMAP specifies the access procedues exercised by network clients to manipulate multipoint calls; it is thus a User-Network Interface (UNI) signalling protocol. We define a multipoint call as a group of multipoint connections. A multipoint connection is a communication channel between two or more clients or endpoints of the network, where all data sent by one client is received by all other clients who have elected to receive. A point-to-point …

Go to article

An Incremental Distributed Algorithm For Computing Biconnected Components, Bala Swaminathan, Kenneth J. Goldman Jan 1994

An Incremental Distributed Algorithm For Computing Biconnected Components, Bala Swaminathan, Kenneth J. Goldman

All Computer Science and Engineering Research

This paper describes a distributed algorithm for computing the biconnected components of a dynamically changing graph. Our algorithm has a worst case communication complexity of O(b + c) messages for an edge insertion and O(b' + c) messages for an edge removal, and a worst case time complexity of O(c) for both operations, where c is the maximum number of biconnected components in any of the connected components during the operation, b is the number of nodes in the biconnected component containing the new edge, and bprime is the number of nodes in the biconnected component in which the update …

Go to article

Morphing Binary Trees, John Hershberger, Subhash Suri Jan 1994

Morphing Binary Trees, John Hershberger, Subhash Suri

All Computer Science and Engineering Research

We investigate the problem of transforming one binary tree into another by rotatoins, subject to certain weight ocnstraints on the nodes of the trees. These constraints arise in the problem of "morphing" one simple polygon to another simple polygon by continuous deformatinos (translations and scalings) that preserve the turn angles and the simplicity of the polygon; the two polygons must have the same sequence of turn angles. Our main theorem is that two arbitrary n-leaf binary trees satisfying our weight constraint can be morphed into each other with O(n log n) rotations. Furthermore, we also present an O(n log n) …

Go to article

Distributed Multimedia Systems Research Prospectus, Jonathan Turner Jan 1994

Distributed Multimedia Systems Research Prospectus, Jonathan Turner

All Computer Science and Engineering Research

Distributed multimedia computing and communiation systems combine computer systems, networks and distributed software to facilitate applications that enable and enhance collaborative work, direct interpersonal communication, remote access to information and real-time presentation of information from a variety of sources. Over the next decade, we expect such systems to become central to the infrastructure of our increasingly information-driven society. This prospectus describes a program of research being pursued within the Computer and Communications Research Center of Washington University and the Departments of Computer Science and Electrical Engineering. This program seeks to promote the creation of effective distributed multimedia systems and develop …

Go to article

An Application-Oriented Error Control Scheme For High Speed Networks, Fengmin Gong, Gurudatta Parulkar Jan 1994

An Application-Oriented Error Control Scheme For High Speed Networks, Fengmin Gong, Gurudatta Parulkar

All Computer Science and Engineering Research

Many new network applications demand interprocess communication (IPC) services that are not supported by existing transport protocol mechanisms. Large bandwidth-delay products of high-speed networks also render the existing control mechanisms such as flow and error control less efficient. In particular, new error control schemes that can provide variable degrees of error recovery according to the applications requirements are needed. This paper presents the design, evaluation, and implementation of an application-oriented error control scheme that is aimed at supporting efficient IPC in high-speed networking environments. Our results show that the proposed error control scheme allows effective control of trade-off between the …

Go to article

Exact Learning Of Discretized Geometric Concepts, Nader H. Bshouty, Paul W. Goldberg, Sally A. Goldman, H. David Mathias Jan 1994

Exact Learning Of Discretized Geometric Concepts, Nader H. Bshouty, Paul W. Goldberg, Sally A. Goldman, H. David Mathias

All Computer Science and Engineering Research

We first present an algorithm that uses membership and equivalence queries to exactly identify a discretized geometric concept defined by the unioin of m axis-parallel boxes in d-dimensional discretized Euclidean space where each coordinate can have n discrete values. This algorithm receives at most md counterexamples and uses time and membership queries polynomial in m and log(n) for any constant d. Furthermore, all equivalence queries can be formulated as the union of O(mdlog(m)) axis-parallel boxes. Next, we show how to extend our algorithm to efficiently learn, from only equivalence queries, any discretized geometric concept generated from any number of halfspaces …

Go to article

Performance Comparison Of Asynchronous Adders, Mark A. Franklin, Tienyo Pan Jan 1994

Performance Comparison Of Asynchronous Adders, Mark A. Franklin, Tienyo Pan

All Computer Science and Engineering Research

In asynchronous systems, average function delays principally govern overall throughput. This paper compares the performance of six adder designs with respect to their average delays. Our results show that asynchronous addres (32 or 64-bits) with a hybrid structure (e.g., carry-select addres) run 20-40% faster than simple ripple-carry addres. Hybrid adders also outperform high-cost, strictly synchronous conditional-sum adders.

Go to article

A Characterization Of The Computational Power Of Rule-Based Visualization, Kenneth C. Cox, Gruia-Catalin Roman Jan 1993

A Characterization Of The Computational Power Of Rule-Based Visualization, Kenneth C. Cox, Gruia-Catalin Roman

All Computer Science and Engineering Research

Declarative visualization is a paradigm in which the process of visualization is treated as a mapping from some domain (typically a program) to an image. One means of declaring such mappings is through the use of rules which specify the relationship between the domain and the image. This paper examines the computational power of such rule-based mappings. Computational power is measure using three separate criteria. The first of these uses the Chomsky hierarchy, in which computational power is treated as string-acceptance; with this criterion we are able to show that certain rule-based models are equivalent in power to Turing machines. …

Go to article

A Fault Tolerant Connectionist Architecture For Construction Of Logic Proofs, Gadi Pinkas Jan 1993

A Fault Tolerant Connectionist Architecture For Construction Of Logic Proofs, Gadi Pinkas

All Computer Science and Engineering Research

This chapter considers the problems of expressing logic and constructing proofs in fault tolerant connectionist networks that are based on energy minimalism. Given a first-order-logic knowledge base and a bound k, a symmetric network is constructed (like a Boltzman machine or a Hopfield network) that searches for a proof for a given query. If a resolution-based proof of length no longer than k exists, then the global minima of the energy function that is associated with the network represent such proofs. If no proof exist then the global minima indicate the lack of a proof. The network that is generated …

Go to article

The Study Of Computer Science Concepts Through Game Play, Benjamin M. Weber Jan 1993

The Study Of Computer Science Concepts Through Game Play, Benjamin M. Weber

All Computer Science and Engineering Research

No abstract provided.

Go to article

Reasoning About Synchrony Illustrated On Three Models Of Concurrency, Gruia-Catalin Roman, Jerome Plun Jan 1993

Reasoning About Synchrony Illustrated On Three Models Of Concurrency, Gruia-Catalin Roman, Jerome Plun

All Computer Science and Engineering Research

This paper presents a model of concurrency (Dynamic Synchrony) whose distinctive feature is a novel formal treatment of synchronization. Synchrony is defined as the coordinated execution of two or more actions. The dynamic aspect comes from the fact that the definition of which actions must be executed synchronously can change freely during the execution of the program. This unique modeling capability comes with a UNITY-stype assertional logic that can be applied to program verification and derivation. This paper shows that the proposed proof logic can be used to verify programs expressed using other models of foncurrency without having to translate …

Go to article

Improving The Speed Of A Distributed Checkpointing Algorithm, Sachin Garg, Kenneth F. Wong Jan 1993

Improving The Speed Of A Distributed Checkpointing Algorithm, Sachin Garg, Kenneth F. Wong

All Computer Science and Engineering Research

This paper shows how Koo and Toueg's distributed checkpointing algorithm can be modified so as to substantially reduce the average message volume. It attempts to avoid O(n{squared}) messages by using dependency knowledge to reduce the number of checkpoint request messages. Lemmas on consistency and termination are also included.

Go to article

Objective Evaluation Of Radiation Treatment Plans, Nilesh L. Jain, Michael G. Kahn Jan 1993

Objective Evaluation Of Radiation Treatment Plans, Nilesh L. Jain, Michael G. Kahn

All Computer Science and Engineering Research

The evaluation of radiation treatment plans involves making trade-offs among doses delivered to the tumor volumes and nearby normal tissues. Evaluating state-of-the-art three-dimensional (3D) plans is a difficult task because of the huge amount of planning data that needs to be deciphered. Multiattribute utility theory provides a methodology for specifying trade-offs and selecting the optimal plan from many competing lans. Using multiattribute utility theory, we are developing a clinically meaningful objective plan-evaluation model for 3D radiation treatment plans. Our model incorporates three of the factors involved in radiation treatment evaluation - treatment preferences of the radiation oncologist, clinical condition of …

Go to article

The Pessimism Behind Optimistic Simulation, George Varghese, Roger D. Chamberlain, William E. Weihl Jan 1993

The Pessimism Behind Optimistic Simulation, George Varghese, Roger D. Chamberlain, William E. Weihl

All Computer Science and Engineering Research

In this paper we make an analogy between the time that storage must be maintained in an optimistic simulation and the blocking time in a conservative simulation. By exploring this analogy, we design two new Global Virtual Time (GVT) protocols for Time Warp systems. The first simple protocol is based on the null message scheme proposed for clock advancement in some conservative approaches; this yields what we call Local Guaranteed Time. Our main contribution is a second new protocol that is inspired by Misra's circulating marker scheme for deadlock recovery in conservative simulations, and appears to have advantages over previous …

Go to article

First Prev Next Last