改进的变步长LMS自适应算法及其应用

在对一些变步长LMS算法分析的基础上,通过改变步长公式中平方根运算范围,提出了一种改进的变步长LMS算法,并进行了稳定性分析.该算法比固定步长算法收敛速度快,稳定性好;与同类型复杂度其他算法相比,具有较快的收敛速度.将该算法应用于自适应噪声对消系统的仿真.计算机仿真证明该算法的理论分析.     

改进的变步长LMSA适应算法蘼其应用

在对一些变步长LMS算法分析的基础上,通过改变步长公式中平方根运算范围,提出了一种改进的变步长LMS算法,并进行了稳定性分析．该算法比固定步长算法收敛速度快,稳定性好;与同类型复杂度其他算法相比,具有较快的收敛速度．将该算法应用于自适应噪声对消系统的仿真,计算机仿真证明该算法的理论分析．     

果蝇优化算法(FOA)步长改进及其多元函数最优化方法

果蝇优化算法(FOA)模拟果蝇群体利用嗅觉和视觉寻找食物的方法来寻找最优值.本文根据算法的特点分析了影响收敛速度的因素,通过变步长方式得到改进的FOA.另外还提出了多元函数最优问题的FOA方法.在对Schaffer函数的全局最优过程中,经过变步长的FOA收敛速度大大提高(与理论最优值的误差以指数速度下降),在计算速度和收敛精度方面都远远优于遗传算法.     

基于分数采样的变步长LMS算法

提出了一种基于分数采样的变步长LMS算法,该算法利用分数采样使之含有更多的信道信息,从而得到更小的稳态误差,利用变步长来加快算法的收敛速度。与常规LMS算法,基于分数间隔的LMS算法相比,新算法具有更快的收敛速度和更小的稳态误差。电话信道的仿真结果,验证了算法的有效性。     

一种变步长LMS自适应滤波算法及分析

讨论了变步长自适应滤波算法,给出了一种新的变步长LMS自适应算法,建立了步长因子μ与迭代次数n及输入信号之间的一种新的非线性函数关系。该算法与传统LMS算法相比,在计算量增加不太大的情况下,具有更快的收敛速度和较小的稳态误差,而且无须去关心步长因子的大小。通过Matlab模拟仿真的结果与理论分析相一致,证实了该算法的优越性。     

有源电力滤波器的箕舌线变步长LMS谐波检测方法

将单相有源电力滤波器（APF）作为研究对象,对其负载非线性电流中的谐波电流检测方法进行了进一步的研究．基于自适应噪声对消技术,将基于箕舌线的变步长最小均方（LMS）算法应用于APF谐波检测,同时将其与定步长LMS算法、基于s函数的变步长LMS算法,以及目前最常用的谐波检测算法之一的离散傅里叶滑窗谐波检测方法进行仿真比较．通过仿真分析,该算法具有较小的运算量、稳态误差小、强跟踪能力等特点,并有较好的收敛速度,解决了收敛速度和稳态误差之间的矛盾．     

基于仿射投影的CMA盲均衡算法

经典CMA算法计算复杂度低,性能稳健,但收敛速度慢.传统的仿射投影算法能够加快算法的收敛速度,但算法复杂度较高点.针对以上问题,该文将仿射投影思想引入到CMA算法,克服了两种算法的缺点,加快了算法的收敛速度,降低了算法的复杂度.通过实验仿真表明,该算法适用于短波信号的盲均衡,且复杂度较低,收敛速度更快.     

基于自适应步长的改进布谷鸟算法

布谷鸟搜索算法在后期搜索过程中存在速度慢、计算精度低等问题,通过引入一种在迭代过程中发现概率和缩放因子自适应策略,对自适应步长布谷鸟搜索算法进行改进,以提高算法的收敛速度和精度.采用4个benchmark测试函数,对基本、自适应步长以及改进的布谷鸟算法进行比较讨论,验证改进算法的有效性.实验结果表明,经改进后的算法具有较好的收敛速度和精度.     

基于模拟退火步长的粒子群算法

为了提高粒子群算法的收敛速度和全局收敛性,本文在标准粒子群算法的基础上作了改进,提出了一种带模拟退火步长的粒子群算法．通过典型函数的测试结果表明新算法比原来算法收敛到最优解的次数多,提出的新算法在全局搜索能力和收敛速度方面有所提高．     

求解蛋白质折叠问题的变步长拟物算法

通过模拟物体间相互作用的物理规律,提出一种求解蛋白质结构预测问题的新型启发式算法。该方法在每一步迭代中,首先利用拟物思想寻找到合理的优化路径,然后在此路径上确定一个有效步长来取得最佳优化效果。若干典型算例的计算结果表明,变步长拟物算法比常规拟物算法具有更好的收敛速度。     

An ensemble approach in converging contents of LMS and KMS

Currently the challenges in e-Learning are converging the learning content from various sources and managing them within e-learning practices. Data mining learning algorithms can be used and the contents can be converged based on the Metadata of the objects. Ensemble methods use multiple learning algorithms and it can be used to converge the Learning Objects from Learning Management Systems (LMS) and Knowledge Objects from Knowledge Management System (KMS). This can increase the performance of the learning system, especially when there is different content available from a variety of models. In this research, Data mining ensemble techniques are used so that an appropriate learning content is delivered to the learner. By converging, the learning content from various sources the Learning system pedagogies can also be revolutionized and a right learning path can be provided to the learners. This research work uses various classification techniques for converging and are evaluated using statistical measures.     

Face-to-face versus computer-mediated discussion of teaching cases: Impacts on preservice teachers’ engagement, critical analyses, and self-efficacy

Two frequently-used discussion protocols were investigated as part of a program to implement teaching cases in undergraduate educational psychology classes designed for preservice teachers. One protocol involved synchronous face-to-face (FTF) discussion of teaching cases, which occurred in class after students had individually completed written case analyses as homework outside of class. The other was asynchronous computer-mediated (CM) discussion taking place outside of class, simultaneous to students’ completion of their written case analyses. Six class offerings of an undergraduate child development course taught by two instructors (three classes by each instructor) were randomly assigned within instructor in a quasi-experimental design to one of the three discussion conditions: FTF, CM, or no discussion of the cases across the semester. Findings indicated that both CM and FTF discussion conditions were associated with positive outcomes relative to the control condition. Both CM and FTF discussion related to higher cognitive–affective engagement with the cases than the control condition; and the CM discussion condition was associated with higher cognitive–affective engagement than FTF discussion. In contrast, FTF discussion, but not CM discussion, was associated with higher-than-control-condition case analysis ability at the end of the semester. Potential explanations for these findings and directions for future research are discussed.     

The impact of electronic media on faculty evaluation

With the proliferation of computer networks and the increased use of Internet‐based applications, many forms of social interactions now take place in an on‐line context through Computer‐Mediated Communication (CMC). Many universities are now reaping the benefits of using CMC applications to collect data on student evaluations of faculty, rather than using paper‐based surveys in Face‐To‐Face (FTF) classroom settings. While the relative merits of CMC versus FTF student evaluations have been researched extensively, there is limited research published about the ways students respond to the questions from either mode of data collection. This paper reports on a research study to analyse the communication differences between student scores from FTF student evaluations and CMC evaluation questions from end of semester evaluations from a university in the Middle East region. In addition to the questions about communication mode differences between two evaluation questions, several demographic variables were measured to determine any interaction effects. The results of our study suggest that the type of communication channel mitigates the responses that students make on CMC evaluations vis‐à‐vis FTF evaluations of faculty. In particular, even though there were significant differences found at the aggregate level between CMC and FTF evaluations, when the course and instructor are controlled for, there were no significant differences reported. In addition, several differences were noted depending on the type and level of the course being studied. Also, we found that students are more likely to express more extreme responses to scale questions in CMC than FTF evaluations. Administrators should consider these potential differences when implementing on‐line evaluation systems.     

An Examination of the Outcomes of a Brief and Innovative Partnership: SJSU and Udacity

In an examination of a brief and innovative partnership, we compared outcomes for two disciplines, Elementary Statistics and General Psychology, across three formats: online as part of the San José State University-Udacity partnership (termed SJSU Plus), face-to-face (FTF), and online in a redesigned course offering. We also examine predictors of student performance in the SJSU Plus courses. The first offerings of the SJSU Plus courses showed poorer performance compared to their FTF and redesigned online equivalents. Redesigned online courses and FTF courses had similar pass rates. SJSU Plus course performance was significantly improved in the second offering of the Elementary Statistics course. More completed assignments in the SJSU Plus courses were associated with higher exam scores and final grades. We conclude that mode of delivery did not contribute significantly to variations in pass rates.     

Course Format and Student Learning Styles: A Comparison of Political Science Courses

ABSTRACT

This study investigates whether differences in learning styles exist between students in online and face-to-face (FTF) sections of political science courses taught by three instructors. Some studies suggest that student preferences regarding online or FTF formats are influenced by their preferred modes of learning. Independent learners, for example, may prefer online courses since they provide individualistic opportunities to study outside of the traditional classroom. This study uses original survey data to assign students one of six learning styles in order to assess whether independent learners are more common in online courses. Our analysis finds no significant differences in independent learners when comparing the two formats. This finding runs counter to studies that argue that independent learners tend to prefer online courses. In fact, the only learning style where we observe a meaningful difference among online and FTF formats is among dependent learners. Contrary to expectation, students enrolled in online versions demonstrated a greater tendency toward dependent modes of learning. Further survey responses suggest that student lifestyle drove course format selection rather than learning style. These findings have important implications for universities that increasingly turn to online courses to address decreasing enrollments and attempt to remedy the high attrition rates associated with those courses.     

What effect does flipping the classroom have on undergraduate student perceptions and grades?

There is a lack of consensus of the effects on student perceptions and performance in flipping the classroom and its possible value over the traditional face-to-face (FTF) classroom approach. This research examines the expectation that flipping an undergraduate, introductory level, information concepts and skills class would benefit student learning and evaluates the results of flipping this type of course. The study showed a significant difference in student perceptions about the course itself between web-enhanced, FTF and hybrid, flipped classrooms, but this may be dependent upon the level of technology integration rather than the approach. In addition, students who were required to work in groups on a weekly basis had more positive responses toward the flipped classroom improving their soft skills than those who did not. Finally, for written content exams it seems the flipped class approach is at least as effective as the traditional classroom in terms of grades. However, it appears teaching software skills may be better with at least some direct instruction over just the use of simulation software.     

CASIO fx-4500P计算器在建筑施工变形监测测边交会法中的应用研究

测边交会法可用于建筑施工变形监测中,传统的方法是钢尺丈量边长交会法,随着电子技术的发展,电子测距仪的应用,特别是全站仪的逐步推广与应用,测边交会法用于建筑施工变形监测又有广泛应用的趋势。但交会点的坐标及误差计算比较复杂,经作者进行测边交会的误差分析的计算结果公式,应用可编程序型计算器fx-4500P及开发的程序,可快速、准确计算测边交会点的坐标和点位精度。     

大学生网络心理咨询可行的理论与技法

网络心理咨询和面询之间存在着身体在场与不在场、人格影响力的弱参与与强参与以及符号应用的全面性与否的差异。以上差异是确定网络心理咨询可行理论的筛选标准的关键。根据筛选标准确定了以＂知＂和＂行＂为主的理论以及以＂情＂为主咨询技法。     

基于压缩感知的电能质量信号重构算法研究

为解决传统电能质量信号在采样时面临的采样率高,采样资源浪费和硬件成本高的问题,压缩感知理论被引入到电能质量信号的采样与重构过程。信号的稀疏表示是压缩感知理论中的关键问题,一般选择正交基作为压缩感知中的稀疏变换基。基于多重扰动的电能质量信号,本文提出了基于不同干扰的电能质量模型来选择不同的信号稀疏变换基的压缩感知重构算法。实验证明与整个信号采用单一DCT变换基或FFT变换基的压缩感知重构算法相比,本文提出的方法具有更好的信号重构性能。     

NBIC会聚技术对促进现代教育技术发展的启示

“NBIC会聚技术”代表着当代科学技术发展的最前沿,本文介绍了“NBIC会聚技术”的背景和含义,提出了NBIC会聚技术对现代教育技术在学科理念、研究领域和专业人才培养等三个方面的启示。