铀矿边坡稳定性的支持向量回归分析

简要介绍了基于统计学习理论的支持向量机回归(SVR)原理,针对边坡稳定性影响因素的复杂性,结合实例运用SVR技术构建了铀矿边坡稳定性的支持向量回归预报模型,并利用网格搜索与留一交叉验证方法(LOOCV)优化模型参数。研究表明,在小样本条件下,SVR预报模型对训练样本的计算值与实测值平均相对误差(MRE)为0.045967%,相对均方误差(MSRE)为0.046371%,拟合值(VOF)为1.999995765,相关系数(R)为0.9984,均比人工神经网络方法的相应指标值要小,说明支持向量回归方法是一种科学有效的矿山边坡稳定性的分析方法。     

Predicting Academic Success: A Longitudinal Study of University Design Students

This article presents a longitudinal study, over eleven years, of the academic progress of a cohort of design students (n = 475) at a major Australian university. The students were from four different spatial design disciplines: architecture, industrial design, interior design, and landscape architecture. The article identifies cognitive variables that may predict future academic success. This research uses regression analysis to identify correlations across a range of variables, in particular exploring the relationship between university entrance scores, subjects studied at school (art and graphics) and academic success as defined by university grade point average and time to completion. This study shows that prior knowledge in art and graphics is not a useful predictor of future academic success, while university entrance scores and first year performance are related to ongoing academic success. These results suggest that for the design disciplines explored here, discipline‐specific prior knowledge is not an important requirement for university entrance.     

The interaction of student and program variables for the purpose of developing a model for predicting graduation from graduate programs over a 10-year period

The purpose of this study was to investigate whether there is some factor or set of factors which will predict the probability of graduation from doctoral programs. This study concentrated on two areas: (a) the factors available to the selection committee when the students apply for admission into graduate programs, and (b) those factors emerging after admission, resulting from the students' meeting program-imposed requirements for graduation. Data were collected from the file of each student admitted to the doctoral program in Educational Administration. The results of the study indicated that program variables are more important than admission variables in predicting graduation. and the technique of Path Analysis is highly recommended to study the problem of student progression through graduate departments.     

盈余项目细分与经营现金流量预测的关系研究

利用1998～2007年提供完整财务报告信息的499家深沪A股上市公司数据,通过对盈余项目的层层分解,发现经营现金流预测模型的预测能力随着项目分解而提高。此外,还发现用现金流量和应计项目衡量时经营活动业绩均不佳的企业,现金流量难以预测,业务较好的企业容易预测。     

在课堂教学中培养学生听力策略之预测能力——以一节听力课为例

听力是语言能力的基础组成部分。听力问题由各种因素引发,听力策略的缺失就是其中一项。本文以一节听力课的设计和实施为例,在课堂教学中培训学生使用听力策略中的预测策略,从而提高听力水平。     

试论《国语》中歌谣谚语的政治功能

歌谣谚语在先秦社会中发挥着重要的文化功能,而其中最重要的就是政治功能。《国语》中谚语的政治功能表现为臣下引谚语以劝谏君主、外交场合的利器、观政知人的依据。《国语》中歌谣的政治功能表现为:抒发讥刺与赞美情绪的议政功能;游说功能;童谣的预言功能。     

谐波分析在旱涝气候预测中的应用

通过对气候要素旱涝的谐波分析建立预测模型，对历史文献资料和现代观测资料进行了分级处理，采用C语言编程，将周期函数(旱涝序列)展开，是把一个比较复杂的周期运动，看成是许多不同频率的简谐振动的叠加。对旱涝气候进行预测并先后成功地预测了1995年、1997年的干旱和2003年的雨涝，效果良好。     

预测技能在英语听力理解中的应用

英语作为必修课一直为人们所重视,尤其近年来更是把听说能力提到了首位。本文则主要是从预测技能的角度来对课堂英语听力理解教学提出几点建议,从听前预测、听时预测和听后推测三个方面出发,并结合实例进行分析,目的在于能够使学生加快对英语听力资料的反应速度,提高学生的听力综合能力。     

应用QSPR方法预测烷烃的临界体积

根据分子拓朴学原理,通过用顶点矩阵和距离矩阵表征分子结构,将基团贡献法和拓朴方法有机地结合在一起,发展一种根据分子结构信息预测烷烃临界体积的新方法,对158种烷烃的预测结果表明,临界体积预测值与实验值的一致性令人满意,平均相对误差1．56％,计算精度优于文献方法。     

大学生英语水平提高的瓶颈及对策

怎样提高英语听力水平一直是困饶大学生的一个难题.特别是英语专业四、六级考试的听力部分,做的不尽如意,许多学生甚至还产生畏难情绪.本文就大学生听力畏难情绪的原因进行了分析,并对怎样提高大学生英语听力水平提出了一些方法和技巧.