7294 research outputs found
Sort by
Laser Short-Pulse Heating of a Three Layer Assembly and the Seebeck Effect
Get PDFLaser short pulse heating of a multi-layer assembly, which consists of different layer properties, results in a non-similar electron and lattice site temperature distributions in the layers. This is because the differences in the amount of energy transfer in each layer despite the fact that each layer is very thin. Consequently, an investigation into the temperature distribution in the electron and lattice subsystems in each layer is essential. In the present study, laser short-pulse heating of a three layer assembly, consisting of Au-Cr-Cu, is examined. The electron and lattice site temperature rise in each layer is predicted using an electron lattice theory approach. Three-dimensional heating situation is accommodated in the model study. The Seebeck coefficient in each layer is computed and compared with the results of the previously derived equation. It is found that the electron temperature distribution varies in each layer and that this variation affects the lattice site temperature distribution. The lattice temperature distribution in the radial direction is not influenced by the diffusion of energy in the radial direction. Abrupt changes in the Seebeck coefficient across chromium and copper layers are observed
Convergence and steady-state analysis of the normalized least mean fourth algorithm
No full textThe normalized least mean-fourth (NLMF) algorithm is presented in this work and shown to have potentially faster convergence. Unlike the LMF algorithm, the convergence behavior of the NLMF algorithm is independent of the input data correlation statistics. Sufficient conditions for the NLMF algorithm convergence in the mean are obtained and an analysis of the steady-state performance is carried out with a new approach. The latter uses the concept of feedback and bypasses the need for working directly with the weight error covariance matrix. Simulation results obtained in a system identification scenario confirms the theoretical predictions on performance of the NLMF algorithm
Infusing Critical Thinking Skill Classification into a Software Engineering Course
Get PDFLife long learning is important to keep oneself up-to-date in ones profession. Due to the rapid evolutionary nature of computer science, life long learning becomes even more important. Equipping the students with critical and creative thinking skills can make learning more effective. Critical and creative thinking skills can be taught either by offering explicit courses on such topics or the important skills can be infused into the contents of various courses in the computer science or computer engineering programs. Teaching critical skills along with the course contents can prove itself more appropriate than only transferring the subject knowledge (course content). Some topics may provide a very natural way to teach a critical thinking skill. This paper describes some of our efforts in infusing the critical thinking skill of classification into a course on Principles of Software Engineering in our undergraduate computer science curriculum
A Neural Network Approach For Estimating Examinees' Proficiency Levels in Computerized Adaptive Testing
Get PDFThis paper studies the potential of using neural network models for estimating examinees' proficiency levels in computerized adaptive testing. Computerized adaptive testing (CAT) has recently become increasingly important to standardized testing. An essential constituent of CAT is the estimation of each examinee proficiency level. Previously, this estimation has been carried out using the maximum likelihood estimator (MLE) or a Bayesian procedure. As being parametric techniques, the quality of estimates strongly depends on some restrictive assumptions. Neural network, with its strong theoretical background and ability to learn and generalize, provides a more flexible non-parametric function approximation and tends to be more efficient in estimation accuracy. It can be used for estimating the proficiency levels of examinees. In this work, several models have been simulated and compared namely multi-layer perceptron (MLP), principal-component analysis (PCA), radial-basis function (RBF) and support-vector machines (SVM). Simulation results reveal that neural-network models are capable of providing good estimates of examinees' proficiecy levels. The accuracy of the classification estimation varies based on the network complexity and the size of the training data set
On optimal firewall rule ordering
Get PDFIn today's online connected world, almost all corporate networks use some form of perimeter firewalls to manage Internet connections and enforce a security policy at the corporate gateway. Although it can considerably enhance network security and protect business-critical information, a firewall with thousands of rules can become a bottleneck for network performance. The primary goal of this paper is to present a new rule order optimizer based on simulated annealing to find optimal configurations that minimize the average number of rule comparisons while preserving precedence relationships among disjoint rules. The proposed approach is evaluated and its effectiveness is compared with another approximate solution under several firewall configurations and policy profiles
Bacterial Growth Classification with Support Vector Machines: A Comparative Study
Get PDFIn this paper, we propose to use support vector machines for classification of bacterial growth and non growth database and modeling the probability= of growth. Unlike artificial neural networks paradigms, support vector machines use the kernel functions and support vectors with maximum margin, which allows a better performance. As a practical application of the new approach, support vector machines were investigated for their quality and accuracy in classifi-cation of growth/no-growth state of a pathogenic Escherichia coli R31 in response to temperature and water activity. A comparison with the most common used statistics, machine learning, and data mining schemes was carried out. The results shows that support vector machines classifier based on the Gaussian RBF Kernel was found to do better than most of logistic regression, K-nearest neighbor, probabilistic networks, and multilayer perceptron classifiers