Visual field mean deviation and relevant factors in 928 Chinese retinitis pigmentosa patients

AIM: To investigate the associations between demographic and clinical factors with the rate of visual field mean derivation(MD) decline in retinitis pigmentosa(RP) patients.METHODS: Correlations of MDs with the visual acuity and retinal pigmentation were analyzed in 928 RP patients. MD decreasing rate in 10 y and potential influences of gender, age, family history and retinal pigmentation on the rate were explored in 201 RP patients. RESULTS: In the 928 patients, average MD and visual acuity were-14.44±8.61 dB and 0.79±0.35 respectively and when MD was lower than-9.18 dB the visual acuity would be below 1.0(20/20). The average MD medium between eyes with or without retinal pigmentation was-14.82 dB. In 123 non-pigmented eyes, the average MD were lower than the medium but in 153 pigmented eyes it was higher than that. In the 201 patients, the average decreasing value of MD in 10 years’ period was-8.01±3.66 dB and the value were correlated to retinal pigmentation but not to gender, age or RP family history. CONCLUSION: The rate of MD decline in RP eyes is significantly related to retinal pigmentation. Our study demonstrates the quantitative rate of MD decline in RP patients and the value of MD could well reflect the severity of RP.

Mean Difference and Mean Deviation of Tukey Lambda Distribution

The purpose of this paper is to broaden the knowledge of mean difference and, in particular, of an important distribution model known as Tukey lambda, which is generally used to choose a model to fit data. We have obtained compact formulas, which are not yet reported in literature, of mean deviation and mean difference related to the said distribution model. These results made it possible to analyze the relationships among variability indexes, namely standard deviation, mean deviation and mean difference, regarding Tukey lambda model.

MEAN SQUARE DEVIATION ANALYSIS OF INTERANNUAL SST VARIABILITY IN TROPICAL PACIFIC AND INDIAN OCEAN

Using the SST data series in tropical ocean (20N ~ 20S, 50E ~ 80W) during 1951 ~ 1997 to calculate its monthly mean square deviation, the work obtains results showing that interannual SST variability of the Pacific is more significant than that of the Indian Ocean, especially near the central and eastern equatorial Pacific (165W～90W, 6N～6S), where it ranges from 2C to 4C. The interannual SST variability is obvious in November and December but small in March and April. The interannual variability of 搘arm pool?SST is not so obvious as that of the eastern equatorial Pacific. However, interannual SST variability of the Indian Ocean ranges from 1C to 2C or so, being smaller than that of the Pacific. In the Indian Ocean, interannual SST variability of the Southern Hemisphere is more obvious than that of the Northern Hemisphere. According to above characteristics of interannual SST variability, the key sectors are determined.

The Mean Deviation from the Median of the Dagum Distribution

The Dagum model is particularly suitable for the analysis of the distributions of economic quantities, such as income, assets and consumption. The purpose of this note is to derive the expression of the mean deviation from the median of the Dagum distribution to study the behavior of the scale and shape parameters in terms of absolute variability and in terms of relative variability.

Spectrum Deviation Method of Identification of Foreshocks or Generalized Foreshocks and Its Applications

Based on previous research work,we present a spectrum deviation method to recognize a foreshock or generalized foreshock in this paper. The criterion to determine whether an event is a foreshock is a wide spectrum for an ordinary event,however,a moderate earthquake with foreshock or generalized foreshock has the characteristics of a narrow frequency band,and it deviates to the low frequency. It may be explained by metastable extension in the rupture source or related area of the main shock or regional fragmentation damage and crack nucleation process. The calculation results of two foreshocks,the M_S4. 7 event which occurred before the Yushu M_S7. 1 earthquake on April 14,2010 and the M_S5. 3 event which occurred before the Yutian M_S7. 3 earthquake on February 12,2014,show that the spectra of foreshocks shift,and they are quite different from the nonforeshock seismic spectrum of equivalent size. Therefore,this result can verify the validity of the spectrum deviation method.

An Efficient Frontier for Sum Deviation JIT Sequencing Problem in Mixed-model Systems via Apportionment

In this paper, the sum deviation just-in-time （JIT） sequencing problem in mixed-model production systems is studied relating with the discrete apportionment problem together with their respective mathematical formulations. The assignment formulation for the first problem is briefly discussed followed by the existence of JIT cyclic sequences. Presenting the concise discussion on divisor methods for the discrete apportionment problem, we have proposed two mean-based divisor functions for this problem claiming that they are better than the existing divisors; hence, we found a better bound for the JIT sequencing problem. The linkage of both the problems is characterized in terms of similar type of objective functions. The problems are shown equivalent via suitable transformations and similar properties. The joint approaches for the two problems are discussed in terms of global and local deviations proposing equitably efficient solution.

最小方差优化初始聚类中心的K-means算法

传统K-means算法随机选取初始聚类中心,容易导致聚类结果不稳定,而优化初始聚类中心的K-means算法需要一定的参数选择,也会使聚类结果缺乏客观性。为此,根据样本空间分布紧密度信息,提出利用最小方差优化初始聚类中心的K-means算法。该算法运用样本空间分布信息,通过计算样本空间分布的方差得到样本紧密度信息,选择方差最小(即紧密度最高)且相距一定距离的样本作为初始聚类中心,实现优化的K-means聚类。在UCI机器学习数据库数据集和含有噪音的人工模拟数据集上的实验结果表明,该算法不仅能得到较好的聚类结果,且聚类结果稳定,对噪音具有较强的免疫性能。

一种基于熵和均方差法综合赋权的K-means算法

在传统的K-means聚类算法基础上,本文提出一种基于熵和均方差法综合赋权的Syn-K-means算法。引入综合权重提高聚类结果的类内相似度,从而提高聚类精度。算法中特征权重的计算基于概率论中数字特征的基本描述方法——均方差和信息论中信息特征的基本度量方法——熵;综合赋权系数的选择采用主观设定法求解。实验结果表明,Syn-K-means算法在聚类精度方面优于标准的K-means算法。

一种基于均方差属性加权的K-means算法

在传统的K-means聚类算法基础上提出了一种基于均方差属性加权的MWS-K-means算法。引入特征权重以提高聚类结果的类内相似度(intra-similarities),从而提高聚类精度。考虑到K-means算法采用误差平方和作为聚类准则函数,而误差平方和与概率论中数字特征的基本描述方法———均方差具有较高相似性,算法中特征权重的计算采用均方差法。根据属性的离散程度对欧氏距离进行加权处理,从而用相对距离代替绝对距离来计算类间相似度。实验结果表明:MWS-K-means算法在聚类精度方面优于标准的K-means算法。

一种高效的基于初始聚类中心优化的K-means算法

为解决传统K-means算法初始质心的随机选取以及聚类过程中每个数据样本到聚类中心距离的重复计算问题,提出了一种高效的基于初始聚类中心优化的K-means算法,采用最小方差优化初始质心,通过存储每次迭代中所有数据点的簇标志和到最近聚类中心的距离并用于下一次迭代,避免了重复计算数据点到每个中心的距离。在UCI数据库中五个不同的数据集上进行了测试,对各个算法在聚类准则函数,运行时间以及迭代次数上进行实验结果比较,表明在不降低聚类性能的前提下,减少了迭代次数,缩短了聚类时间,证明了改进算法的有效性和高效性。

基于K-means聚类十字线的套印偏差检测方法研究

目的研究并改进基于十字线套准标识符的印刷品套印偏差检测算法,以提升偏差量检测精度。方法使用K-means聚类算法对分色后得到的C,M,Y,K图像分别进行聚类处理,然后分割图像并提取出各色前景像素,进而使用C,M,Y图像中的前景十字线分别减去K图像中的十字线,以消除四色重叠区域的干扰,最后提取十字线中心并计算偏差量。结果使用15幅印刷品图像样张进行实验,改进的基于K-means聚类十字线的套印偏差检测算法的检测结果,比大津法更接近人工测量值,误差在±0.04mm以内。结论基于K-means聚类的十字线分割法对前景十字线的提取,较大津法抗叠印干扰性更强。

基于标准偏移量的学生成绩K-means聚类分析算法研究

针对利用距离差聚类分析算法分析学生学科成绩不够准确的问题,提出了利用标准差计算标准偏移量构建目标函数的K-means聚类分析算法,给出了初始聚类中心选取办法和算法的描述及处理流程;实验结果分析可得,利用标准偏移量构建的学生成绩K-means聚类分析算法,符合对学生成绩按学科类别聚类分析的特性要求;该算法能够较好的实现学科成绩高度关联属性的聚类分析结果。

基于k-means聚类的智能电表异常动态数据检测方法

为缩小常规电表电量数据与异常测量数据之间的补偿误差水平,提出基于k-means聚类的智能电表异常动态数据检测方法。通过电量负荷异常数据预处理的方式,确定具体的密度偏差值结果,再根据聚类k值条件,提取基于k-means聚类的异常电量数据特征。在此基础上,设定电表数据异常值,根据异常数据清洗原理,得到最终的动态检测特征值计算结果,实现基于k-means聚类智能电表异常动态数据检测方法的设计与应用。实验结果表明,对于常规电表电量数据与异常测量数据而言,k-means聚类检测方法的误差百分比数值较低,能够有效缩小常规数据与异常数据之间的补偿误差水平。

基于改进K-means聚类算法在电力客户价值分群的应用

针对电力客户特点实行不同的营销策略和提供差异化服务,就需要对电力客户做出准确的分群。传统K-means聚类算法对数据分布均匀的类似球形的数据集聚类效果比较好,一旦数据集分布密度不均衡,类簇大小差异明显时,传统K-means算法容易使稀疏的大类簇被高密度小类簇瓜分,导致电力客户分群正确率下降。论文基于电力客户数据分布不均衡的特点,采用了一种改进的K-means聚类算法。改进的K-means算法提出一个新的加权聚类准则,并根据该准则修改了聚类迭代过程。文章最后在对电力客户数据的分群聚类结果表明,改进的K-means聚类算法的分群聚类效果中各个群类的紧凑性得到有效提高,误分情况明显改善。

一种加权K-means聚类算法及其应用

K-means聚类方法是比较经典的聚类分析方法之一,因执行效率快,算法简单等优点在许多领域得到广泛应用.传统的K-means聚类算法没有考虑各指标的重要性,实际上,指标的重要性是有差异的,应当区别对待.基于此,改进传统的K-means聚类方法,提出了加权K-means聚类算法.为验证改进算法的优越性,以R软件自带的鸢尾花卉数据集iris作为实验数据,设定了三个实验进行对比分析,实验结果表明加权K-means聚类算法收敛速度较快,分类的结果更接近原始分类,效果更好.为进一步说明加权K-means聚类算法的实用性,以103所高校部分指标为实验数据用该算法进行了聚类分析.

基于互信息量均方差提取关键帧的激光视频图像检索研究

为保证激光视频图像检索结果中不存在重复性冗余图像,提出了基于互信息量均方差提取关键帧的激光视频图像检索方法。基于互信息量均方差的关键帧提取方法,以激光视频图像颜色的互信息量均方差最大化,为激光视频图像关键帧的聚类中心设置标准,以此聚类提取不重复的视频图像关键帧;通过基于关键帧的激光视频图像检索方法,将所提取关键帧作为激光视频图像检索的核心判断内容,提取与所需图像关键帧相似度显著的激光视频图像,完成激光视频图像检索。实验结果显示:此方法使用后,提取的激光视频图像关键帧冗余度仅有0.01,激光视频图像检索结果的MAP指标测试值高达0.98,检索结果中不存在重复性冗余图像。

基于Kmeans和TOPSIS的电力系统黑启动方案分类评估

传统黑启动方案评估法因步骤繁琐,计算复杂,不仅会增加电力系统调度人员的决策时间,导致电力系统各断电机组不能迅速恢复,还会给社会带来巨大的经济损失和负面影响。本文将视野从决策学领域转向数据挖掘领域,提出一种基于Kmeans和TOPSIS的黑启动方案分类模型。并采用广东电网黑启动数据进行了验证,并利用已有的平均绝对偏差比较策略,对文章所提模型的方案排序与现有权威黑启动方案评估下的排序结果进行量化比较分析,证明了此模型的科学、有效性。

控制图在车用催化器中贵金属测定中的应用

研究建立一种应用于催化器中贵金属测定过程监控的均值-标准差质控图,以确保实验室日常分析结果的准确、可靠。采用标准HJ 509-2009中电感耦合等离子体质谱法测定催化转化器中的铑、钯、铂等3种贵金属元素的含量,收集近两年的质控样的60组数据,进行了正态分布检验,计算了各元素含量结果的平均值、标准偏差,绘制了均值控制图和标准差控制图。结果表明,质控样中铑、钯、铂贵金属元素的数据结果均在均值控制图、标准差控制图的上下限范围内,表明实验室在该段期间内利用电感耦合等离子体质谱法对催化转化器样品中的铑、钯、铂的测定结果的准确可靠。均值-标准差控制图可以有效地观察整个贵金属元素测定过程中的数据变化趋势,有助于判断检测过程是否受控,以保证日常的报告检测结果符合实验室报出的要求,确保结果的准确可靠。

嫦娥4/5号月球探测器VLBI观测数据质量分析

在中国探月工程(China’s Lunar Exploration Program,CLEP)中,甚长基线干涉测量技术(very long baseline interferometry,VLBI)具有重要作用。为获得月球探测器的准确定位、实现精确的测定轨功能并以此在探测器飞行过程中完成轨道修正,需要利用VLBI测站对月球探测器进行实时观测。首先以均值(arithmetic mean,AM)以及标准差(standard deviation,STD)作为主要的精度指标,对来源于上海天文台的嫦娥4号探测器、中继星以及嫦娥5号探测器的VLBI实测数据进行总观测时段的数据稳定性分析,并基于嫦娥4号探测器同时段S、X双频段观测,着重进行该探测器不同频率各项数据的质量分析;然后对探测器各工作时段分别进行时延、时延率测量形式误差以及残差分析,了解各时段探测器运动情况与数据变化的相关性。结果表明,3类探测器中,嫦娥4号的实测精度最高,达到1 ns量级;嫦娥5号的观测稳定性优于其他两类探测器;不同地面站对探测器的观测精度有所差异,并且在探测器不同的工作时段,实测数据表现出不同特征。

基于卷积神经网络的电力市场短期售电量预测方法

电力市场短期售电量预测的精度对优化用电结构以及提高供电可靠性具有重要意义,传统短期售电量预测方法没有考虑偏差电量考核影响、用电行为差异导致电预测精度低,提出基于卷积神经网络的电力市场短期售电量预测方法,首先根据用户的用电负荷率进行分类,获取不同行业的用电特征和需求模式,然后考虑正负偏差电量的影响,设计基于CNN-ResNet的短期售电量预测方法,通过实验分析表明,该方法能够有效提高多因素影响下售电量预测的准确率。