In this paper we consider a method for finding several eigenvalues and corresponding eigenvectors of large-scale generalized eigenvalue problems. In this method, a small matrix pencil that has only the desired eigenva...
详细信息
This paper presents a view of ABET engineering accreditor training from the perspective of an engineering professor from small comprehensive college who is helping prepare a computerengineeringprogram for accreditat...
详细信息
This paper presents a view of ABET engineering accreditor training from the perspective of an engineering professor from small comprehensive college who is helping prepare a computerengineeringprogram for accreditation. The foundation of the paper is based on the author's observations while participating in an ABET accreditor training course at an ASEE conference where 95% of the participants were ABET accreditors. A comparison of ABET evaluation standards as they have changed from the prescribed requirements of the past to the more self-determined measurements of the newer ABET standards yields much insight into the accreditation process. A plan for organizing required documentation, resources, and people is included. The degree to which an engineeringprogram has matured to an "ABET-ready" status must be established to spare a department and college from the financial and emotional costs of planning, funding, and facilitating an ABET accreditation review before a program is ready. Such specifics as having clear objectives, verifiable outcomes, and a fully developed comprehensive curriculum are also discussed. Principles such as self-evaluation, continuous improvement, and life-long learning are critical.
Prototype supercomputer functionality can be verified by comparing simulated hardware execution with actual hardware test-program runs where each successive test-program run includes randomly changing machine-states, ...
详细信息
Prototype supercomputer functionality can be verified by comparing simulated hardware execution with actual hardware test-program runs where each successive test-program run includes randomly changing machine-states, operating scenarios, and data. Increased verification is achieved through repeated program execution. In both multi-processor and vector-register systems, a "controlled randomness" can be used to verify the functionality of simultaneously executing processors or functional units. This paper discusses the selection and combining of random number generators such that a "degree-of-randomness" between successive or parallel program runs is controlled. This allows computer engineers to simulate the execution of actual software (application or system-level) in which successive or parallel program runs may or may not involve uncorrelated tasks. Additionally, random number generators are selected to maximize execution speed and cycle-length, ensure reproducibility, and when desired, best produce a random source of numbers (i.e., to better approximate an independent, identically-distributed source). Generators can also be chosen for ease of implementation, the ability to run backwards, and the ability to split the generator's cycle into uncorrelated segments. "Backward multipliers" to allow generators to be run in reverse can also be easily found for some types of generators;reversibility is critical for functional verification so that code execution can be traced backwards to find scenarios that led to detected hardware failures. When generators are carefully selected and combined, the verification process can be optimized. By using this methodology, functional verification of SMP, MPP and vector-register supercomputers can be achieved.
We present a functional and structural didactic simulator of Cache Memory Systems developed at the Pontifical Catholic University of Minas Gerais, Brazil. The development occurred during the undergraduate computer Arc...
详细信息
Lack of reliable convergence of training algorithms due to flat regions in the quadratic error curve is a most serious problem in neural networks. A new type of training algorithm presented here, collectively called t...
详细信息
Lack of reliable convergence of training algorithms due to flat regions in the quadratic error curve is a most serious problem in neural networks. A new type of training algorithm presented here, collectively called the Glide Algorithm with Tunneling, deals successfully with this problem. This (type of) algorithm converges reliably in several hard test problems and appears to be applicable to general feed-forward, multilevel perceptrons. Tests show fast convergence 100% of the time for Parity-4, character recognition, and Two-Spiral problems. For comparison, Gradient Descent with or without momentum only converges at a 40%-80% rate on the same problems using comparable parameters. Additionally, the average CPU time used by the Glide Algorithm with Tunneling can be as little as 1/12 of that which is used by regular Gradient Descent. The new algorithm's fast, reliable convergence and robustness to parameter variations makes it attractive for general, practical applications.
The wide availability of mobile devices equipped with wireless communication devices paves the way for building highly dynamic communities of ad-hoc networks. For an asynchronous service discovery in this environment,...
详细信息
Restructuring in electric power sectors took a significant step backward in the summer of 2000 when wholesale and retail markets in California experienced staggering price increases that continued through June of 2001...
详细信息
The development of computer outreach opportunities for learning (COOL) project for secondary schools to provide computerprogramming education is discussed. The COOL project is a software development curriculum module...
详细信息
The development of computer outreach opportunities for learning (COOL) project for secondary schools to provide computerprogramming education is discussed. The COOL project is a software development curriculum modules pilot program for secondary teachers to provide students with a better understanding of the software development field. The project also helps teachers in dispeling misconceptions and increasing the number and diversity of the students continuing their studies in preparation for software development careers. The COOL project has a freshman component fro incoming ASU students, consisting of placement test on computer skills expected for the introduction to engineering design course and the beginning course of Java.
Block coders are among the most common compression tools available for still images and video sequences. Their low computational complexity along with their good performance make them a popular choice for compression ...
详细信息
Block coders are among the most common compression tools available for still images and video sequences. Their low computational complexity along with their good performance make them a popular choice for compression of natural images. Yet, at low bitrates, block coders introduce visually annoying artifacts into the image. One approach that alleviates this problem is to downsample the image, apply the coding algorithm, and interpolate back to the original resolution. In this paper, we consider the use of optimal decimation and interpolation filters in this scheme. We first consider only optimization of the interpolation filter, by formulating the problem as least-squares minimization. We then consider the joint optimization over both the decimation and the interpolation filters, using the Variable Projection method. The experimental results presented clearly exhibit a significant improvement over other approaches.
暂无评论