A new learning method using prior information of neural networks

In this paper, we present a new learning method using prior information for three-layer neural networks. Usually when neural networks are used for identification of systems, all of their weights are trained independently, without considering interrelated weights values. Thus, the training results are usually not good. The reason for this in that each parameter has its influence on others during learning. To overcome this problem, we first give an exact mathematical equation that describes the relation between weight values given a set of data conveying prior information. The we present a new learning method that trains part of the weights and calculates the others using these exact mathematical equations. This method often a priori keeps the given mathematical structure exactly the same during learning; in other words, training is done so that the network follows a predetermined trajectory. Numerical computer simulation results are provided to support this approach. This work was presented, in part, at the Fourth International Symposium on Artificial Life and Robotics, Oita, Japan, January 19–22, 1999.     

混合现实虚拟智能教室的方案设计

为了实现在网络学习中,教师和学生能够有效地支持师生实时课堂交互,针对远程教育停留在Web课件浏览以及教师单向讲授的形式和阶段上,讲授形式和画面单一呆板,师生不能充分交流的问题,在分析网络教育存在的缺陷,以及思考网络学习中必须关注的几个问题的基础上,提出了一种新的远程教学互动及演示方式———混合现实虚拟智能教室系统.该系统是一个依靠交互空间技术和电视演播技术而开发的网络教学环境,在利用该系统进行教学时,网络学生可以用不同的接入方式加入该虚拟教室课堂,不仅可以通过多变的虚拟背景下教师的讲课视频接受到更多的感性认识;还可以通过语音、视频以及文本与教师进行直接对话,大大加强了网络学习过程中师生的沟通交流,打破了网络学习时学习者情感的缺失,从而使得教学效果能够加强.     

基于二叉级联结构的并行极速学习机算法

为解决因庞大的矩阵存储和计算,ELM（Extreme Learning Machines）难以应用到大规模、高维数据集的问题,提出一种基于＂分而治之＂策略的并行极速学习机算法。该算法利用二叉级联结构,将大规模数据集分派到多个计算节点上,并行地更新单隐层前馈网络的输出权值,且能有限步地单调收敛到最小二乘解。实验结果表明,该算法不仅泛化性能优异,并且具有非常高的加速比和并行效率。     

基于混合熵和L_1范数的遥感图像分类

针对遥感图像数据具有的高维数、非线性以及海量无标记样本的特性,提出了一种基于混合熵和L1范数的概率型最小二乘支持向量机分类方法.将准熵和熵差分融合,构造一种混合熵用以从海量无标记样本集中选出最有＂价值＂的待标记样本;基于L1范数距离度量,进一步从待标记样本集中筛选出孤立点和冗余点加以剔除;基于初始已标记样本以及筛选得到的样本,训练得到概率型最小二乘支持向量机.对反射光学系统的成像光谱仪（ROSIS）高光谱遥感图像进行了分类实验.结果表明：所提分类方法的总精度和Kappa系数分别达到了89.90%和0.868 5,能够以较少的训练样本得到较高的分类精度,其更适于处理遥感图像分类问题.     

学习型军代室建设中的领导方式问题研究

利用学习型组织理论分析学习型军代室建设过程,寻找在军代室建设的各个阶段军代室领导应该采用的有效领导方式。结合军代室工作实际,运用塔克曼团队建设的阶段理论,提出了学习型军代室建设的5个阶段,即形成阶段、震荡阶段、规范阶段、执行阶段和创新阶段。结合团队建设各个阶段的特征,分析了军代室领导应当承担的任务和充当的角色,指出在学习型军代室建设的不同阶段,军代室领导应分别以榜样者、辅导者、参与者、授权者和观察员5种不同的角色出现。     

关于加强和改进基层党委中心组学习的几点思考

党委中心组学习是本级党委组织本级领导班子成员开展以党的理论和路线方针政策为主要内容的有目的、有意识、有创新的学习活动。对新时期、新形势下的基层党委中心组学习的内涵、地位、作用和基本经验做了分析和思考,并对如何改进当前基层党委中心组学习取得实效提出看法和建议。     

军校新生学员的转型问题探析

为使军校新生学员尽快融入军校的知识学习和军事训练之中,归纳了军校新生学员因环境变化而带来的思想、生活、自主学习、心理和人际交流等5个方面＂不适应＂的主要表现,分析了产生这些＂不适应＂表现的家庭、社会教育和新生自身3种主要原因,提出了加强思想教育和心理适应教育、确立学习奋斗目标、改进学习方法、教员注意教法上的衔接、营造和谐的学习生活环境、发挥队干部和新生干部作用等主要措施.     

军校营造良好学风的几点思考

针对当前军校学风建设存在的问题,从加强思想教育、完善教学管理制度、发挥队干部良好的导向作用、加强教风建设及落实教学管理制度等方面提出了军队院校营造良好学风的对策.     

基于DPI和机器学习的网络流量分类方法

网络流量分类是实现网络管理的重要技术之一,但是单一的基于DPI或是机器学习的分类方法分类精确度低.提出了一种基于DPI和机器学习相结合的网络流量分类方法.该方法采用DPI检测已知特征的网络流量,利用机器学习方法辅助分析未知特征以及加密的网络流.实验表明该方法能够提高网络流量分类的精确度.     

Exploring quadrilaterals in a small group computing environment

Though cooperative learning has been a topic of considerable interest in educational research, there has been little study specific to learning in the mathematics content area of geometry. This paper seeks to address that gap through a design experiment featuring a novel small-group computing environment for supporting student learning about quadrilaterals. In this design, each student controls a unique point in a shared geometric space, and those points are linked such that a group of four students collectively forms a quadrilateral. We first present results from pre- and post-measures to show how the students learned from the activities and developed in terms of geometric reasoning. We then present three episodes, elaborated with the notion of appropriation, to explain how students took up ways of using the technological tools and of talking about geometric concepts from one another in the interactive environment. Our study found that students achieved learning gains in this novel environment, that the environment provided rich opportunities for peer interaction around geometric objects, and that student learning opportunities and interactions were characterized by processes of appropriating ways of talking about and using software features.