基于岭估计的岭马田系统在复共线性数据中的应用

在多变量模式识别领域,变量间经常会存在复共线性,复共线性不仅会影响参数估计的效果,也会使变量的敏感性出现显著异常.马田系统是以马氏距离作为测量尺度的多变量模式识别方法,复共线性会通过马氏距离影响马田系统变量筛选的效果和判别的准确率.基于岭估计提出了一种新的测量尺度—岭马氏距离,利用岭迹法确定岭参数,将其引入马田系统使得马田系统对病态数据具有更好的耐受性.通过案例验证了岭马氏距离可以很好的克服复共线性,并提高马田系统的判别准确率.     

基于多变量控制图的马田系统优化研究

马田系统(MTS)是一种多元模式识别方法,它首先通过正常样本来建立基准空间,再利用正交表和信噪比来筛选有效变量,最后通过马氏距离来进行分类、诊断和预测.当建立基准空间的正常样本中掺杂少数异常点时,MTS的性能必然会受到影响.根据多变量控制图原理对建立基准空间样品的适合性进行判别,将在控制线外的样品点删除后建立新的基准空间,并通过UCI数据集进行可行性分析及分类效果比较,结果显示:经多变量控制图优化后的MTS,其性能得到显著提高.     

基于粗糙集的马田系统改进及其在分类中的应用

马田系统是由日本著名质量工程学家田口玄一提出的一种模式识别方法,它将正交试验设计、信噪比与马氏距离进行集成,筛选有效特征变量,对待测群体进行诊断、评价和预测.马田系统利用正交表和信噪比筛选特征变量可能存在不足之处,而粗糙集是处理不完善、不确定数据等不完全信息并能进行属性约简的有效方法,引入粗糙集筛选有效特征变量以改进马田系统.癌细胞的及早发现有助于乳腺癌的早期预防和及时治疗,以乳腺癌细胞的分类检测为背景,选取UCI数据库中600个细胞作为研究样本,使用改进马田系统方法区分正常细胞和乳腺癌细胞,并将其分类效果与经典马田系统相比较.结果表明,基于粗糙集的改进马田系统对乳腺癌细胞的分类正确率高于经典马田系统,粗糙集方法大大减少了特征变量个数,可简化数据的收集工作,为医疗上乳腺癌疾病的早期诊断及其他实际分类工作提供技术参考.     

采用优化模型指标筛选的马田系统综合评价方法研究

传统的马田系统主要用于分类与诊断.将马田系统作为一种综合评价方法进行研究,分别研究了有基准空间和无基准空间两种情形下的马田系统综合评价方法及步骤.针对传统马田系统变量筛选存在的缺陷,构建多目标规划模型进行评价指标筛选,采用遗传算法求解模型.通过两个实际案例,将马田系统综合评价方法与一些常用的综合评价方法对比研究,结果表明,马田系统可以筛选评价指标和避免指标赋权问题,是一种实用且有效的综合评价方法.     

基于GBMTS算法的不平衡数据分类研究

解决不平衡数据分类问题,在现实中有着深远的意义。马田系统利用单一的正常类别构建基准空间和测量基准尺度,并由此建立数据分类模型,十分适合不平衡数据分类问题的处理。本文以传统马田系统方法为基础,结合信噪比及F-value、G-mean等分类精度,建立了基于遗传算法的基准空间优化模型,同时运用Bagging集成化算法,构造了改进马田系统模型算法GBMTS。通过对不同分类方法及相关数据集的实验分析,表明:GBMTS算法较其他分类算法,更能够有效的处理不平衡数据的分类问题。     

基于改进阈值的马田系统在医疗智能诊断中的应用

人工智能与医疗数据的融合可以加速疾病诊断过程,提高诊断精度,挖掘诊断过程关键指标,改善医疗工作流程.以医疗决策者偏好为前提,提出基于二次损失函数改进阈值的马田系统算法.通过正交表与信噪比对指标进行优化,降低模型复杂度;通过对阈值的改进,提高模型的灵敏度,满足医疗工作者的决策偏好.将此方法应用于UCI乳腺癌医学数据及三甲医院哮喘临床数据,并与其他改进阈值的马田系统算法及智能算法对比分析.结果表明,改进阈值的马田系统算法识别灵敏度高,简化诊断指标,训练耗时少,是一种更为有效的医疗智能诊断方法.     

基于MTALAB的神经网络判别模型

判别分析是判别样品所属类型的一种统计方法.利用M ATLAB提供的神经网络工具箱为基础,设计了一个三层BP神经网络判别模型,提出了一种进行判别分析的新方法,实例表明,利用BP神经网络建立的判别模型是进行判别分析的有效方法.是对研究分类问题的方法的扩充.     

基于判别超图和非负矩阵分解的人脸识别方法

非负矩阵分解是一种流行的数据表示方法,已广泛应用于图像处理和模式识别等问题.但是非负矩阵分解忽略了数据的几何结构. 而现有的基于简单图的学习方法只考虑了图像的成对信息,并且对计算相似度时的参数选择非常敏感. 超图学习方法可以有效地解决这些问题. 超图利用超边将多个顶点相连接用以表示图像的高维结构信息. 然而, 现有的大部分超图学习方法都是无判别的学习方法.为了提高识别效果, 提出了基于具有判别信息的超图和非负矩阵分解方法的新模型, 利用交替方向法进行迭代求解新模型, 并结合最近邻方法进行人脸识别. 在几个常用标准人脸图像数据库上进行实验, 实验结果表明提出的方法是有效的.     

基于控制图的马田系统马氏空间生成机理研究

马田系统是以马氏距离为测量尺度,通过选取正常样本构建马氏空间,对多元系统进行诊断和预测的分类技术。马氏距离对样本数据的变化非常敏感,因此用于构建马氏空间的正常样本的数据质量直接影响到分类的准确率。实际应用中正常样本的选取大多依据主观经验判断,缺乏客观规范的选择机制。本文提出基于控制图的马氏空间生成机理,先由专家选取的正常样本构建初始马氏空间,再以每个正常样品在初始马氏空间和对应的缩减马氏空间上的马氏距离增量作为新的测量尺度,以此建立单值控制图,利用控制图稳定性判定规则剔除异常数据,从而得到稳定状态的马氏空间。实验分析结果表明该方法的有效性且提高了马田系统分类的准确率。     

基于多元密度核估计的Bayes判别法

判别分析是根据所研究个体的观测指标来推断该个体所属类型的一种统计方法,但一般的Bayes判别法都须假设各个判别变量服从正太分布,且经常出现回代误判率不为零的问题.首次将多元密度核估计应用到Bayes判别法中,消除了判别变量要服从正太分布这个条件约束,并应用建立的最优窗宽调整算法,将回代误判率调整为零,最后对实例进行分析看出结论应用效果较好.     

Affine Mendelsohn triple systems and the Eisenstein integers

We define a Mendelsohn triple system (MTS) of order coprime with 3, and having multiplication affine over an abelian group, to be affine, nonramified. By exhibiting a one‐to‐one correspondence between isomorphism classes of affine MTS and those of modules over the Eisenstein integers, we solve the isomorphism problem for affine, nonramified MTS and enumerate these isomorphism classes (extending the work of Donovan, Griggs, McCourt, Opr?al, and Stanovský). As a consequence, all entropic MTSs of order coprime with 3 and distributive MTS of order coprime with 3 are classified. Partial results on the isomorphism problem for affine MTS with order divisible by 3 are given, and a complete classification is conjectured. We also prove that for any affine MTS, the qualities of being nonramified, pure, and self‐orthogonal are equivalent.     

多元系统马氏田口方法的诊断与分析研究

马氏田口方法是研究多元系统的一种新方法,主要用于建立和优化多元系统的测量表,以便采用优化后的测量表进行诊断/预测。本文基于马氏田口方法四大基本步骤和MYT正交分解对马氏田口方法进行了拓展研究。首先,介绍了马氏田口方法的四大基本步骤;接着介绍了MYT正交分解法,并将MYT正交分解法应用于多元系统的马氏田口方法异常值潜在原因分析;最后,利用拓展后的马氏田口方法对某医院血粘度诊断系统进行了优化,同时对利用优化后的测量表进行诊断时发现的异常值进行了潜在原因分析,取得了良好的效果。     

A collocation method to fit geophysical survey data

A self adaptive collocation method has been applied to the quantitative interpretation of geophysical survey data. Examples of interpretation of vertical electrical soundings (VES) and magneto-telluric soundings (MTS) have been chosen for this work due to the high nonlinearity of the response functions and to the high degree of correlation of the parameters involved. In both cases horizontally isotropic multi-layered models have been adopted.For comparison two different computing algorithms have been considered. In the first, using a generalized inverse formulation a new different decomposition into characteristic values has been used. In the second the Marquardt procedure has been modified in an original way to give a faster convergence using a strategy based on the contemporary adjustment of the step length and the direction of the regression. The confidence limits of the results have been evaluated.Results are given for interpretation of both MTS and VES curves using the two different algorithms. Model data and field data have been considered for an effective comparison of speed of convergence and accuracy.     

MTS法用于上市公司财务质量评估初探

本文应用MTS法(MahananobisTaguchiSystem)于我国股票市场,以评估上市公司的财务质量。通过综合分析深圳和上海1000余家上市公司2000年年度财务报表的15项财务比率发现,利用MTS法提供的技术可以较有效识别存在财务质量以及财务造假问题的公司。     

A modified tabu search algorithm for cost-based job shop problem

In this paper, a cost-based job shop problem (JIT-JSP) is proposed to model the multi-order processing procedure in a just-in-time (JIT) environment. The objective of JIT-JSP is to minimize three costs: work-in-process holding cost of half-finished orders, inventory holding cost of finished orders and backorder cost of unfulfilled orders. A modified tabu search (MTS) method is developed to improve the schedule quality by searching the neighbourhood of a feasible schedule iteratively. The MTS method is comprised of three components that help to ensure a more effective searching procedure: neighbourhood structure, memory structure and filter structure. Computational results show that the MTS method significantly improves the initial schedule generated by an arbitrarily selected dispatching rule.     

IPA Derivatives for Make-to-Stock Production-Inventory Systems with Backorders

A single-stage Make-to-Stock (MTS) production-inventory system consists of a production facility coupled to an inventory facility, and is subject to a policy that aims to maintain a prescribed inventory level (called base stock) by modulating production capacity. This paper considers a class of single-stage, single-product MTS systems with backorders, driven by random demand and production capacity, and subject to a continuous-review base-stock policy. A model from this class is formulated as a stochastic fluid model (SFM), where all flows are described by stochastic rate processes with piecewise constant sample paths, subject to very mild regularity assumptions that merely preclude accumulation points of jumps with probability 1. Other than that, the MTS model in SFM setting is nonparametric in that it assumes no specific form for the underlying probability law, and as such is quite general. The paper proceeds to derive formulas for the (stochastic) IPA (Infinitesimal Perturbation Analysis) derivatives of the sample-path time averages of the inventory level and backorders level with respect to the base-stock level and a parameter of the production rate. These formulas are comprehensive in that they are exhibited for any initial condition of the system, and include right and left derivatives (when they do not coincide). The derivatives derived are then shown to be unbiased and their formulas are seen to be amenable to fast computation. The generality of the model and comprehensiveness of the IPA derivative formulas hold out the promise of gradient-based applications. More specifically, since the base-stock level and production rate are the key control parameters of MTS systems, the results provide the theoretical underpinnings for optimizing the design of MTS systems and for devising prospective on-line adaptive control algorithms that employ IPA derivatives. The paper concludes with a discussion of those issues.     

Management of the risk of backorders in a MTO–ATO/MTS context under imperfect requirements

Production systems are often classified according to the way production is released, e.g. make-to-stock (MTS), make-to-order (MTO), assembly-to-order (ATO) or engineer-to-order (ETO). The choice of a type of production depends on the decoupling point between customer and supplier. In some supply chains, like in the aeronautical sector, a customer may work according to a MTO process (since his product is highly specific) while his supplier works with a MTS process (since he delivers variants of standards components). This situation sets specific problems that are seldom considered in the literature, especially when collaboration between actors is required for an efficient management of the supply chain, which is the case when uncertainties are present. In this paper, we propose a method based on fuzzy modelling allowing a customer to choose a plan taking into account the uncertainty on his requirements when he works in MTO–ATO while his supplier is in MTS.