Error rates in quadratic discrimination with constraints on the covariance matrices

In multivariate discrimination of several normal populations, the optimal classification procedure is based on quadratic discriminant functions. We compare expected error rates of the quadratic classification procedure if the covariance matrices are estimated under the following four models: (i) arbitrary covariance matrices, (ii) common principal components, (iii) proportional covariance matrices, and (iv) identical covariance matrices. Using Monte Carlo simulation to estimate expected error rates, we study the performance of the four discrimination procedures for five different parameter setups corresponding to standard situations that have been used in the literature. The procedures are examined for sample sizes ranging from 10 to 60, and for two to four groups. Our results quantify the extent to which a parsimonious method reduces error rates, and demonstrate that choosing a simple method of discrimination is often beneficial even if the underlying model assumptions are wrong.The authors wish to thank the editor and three referees for their helpful comments on the first draft of this article. M. J. Schmid supported by grants no. 2.724-0.85 and 2.038-0.86 of the Swiss National Science Foundation.     

<Emphasis Type="Italic">Naja siamensis</Emphasis>, a cryptic species of venomous snake revealed by mtDNA sequencing

Because of possible variation in venom composition, an understanding of venomous snake systematics is of great importance for the optimization of antivenom treatment of snakebite patients. Intraspecific variation in the morphology of many venomous snakes complicates the definition and indentification of some species when allopatric populations are involved. Selectively neutral or near-neutral mtDNA sequences can reveal evolutionary relationships obscured by ecogenetically-caused morphological variation. We use comparative sequencing of the cytochrome oxidase subunit 1 gene to reveal the existence of a widespread, cryptic species of spiting cobra from southeast Asia. This species,Naja siamensis, is widely sympatric with other Asiatic cobra species. This may be of considerable medical significance, and calls for further research into venom composition in Asiatic cobras.     

Variability of sterol utilization in stored-products insects

A comparison of sterol utilization by 3 stored-products insects revealed very different capabilities. The fluor beetle,Tribolium castaneum dealkylates and converts dietary sitosterol to about equal amounts of cholesterol (43.7%) and 7-dehydrocholeterol (39.8%), whereas another flour beetle,Tenebrio molitor, produces considerably less 7-dehydrocholesterol (16.8%) and relatively more cholesterol (66.7%) from sitosterol. The lepidopteran,Plodia interpunctella, utilized dietary sterol very similar to plant-feeding lepidoptera, producing primarily cholesterol (86.5%) from sitosterol.     

用于VLSI的新型介质膜界面陷阱的特征

采用雪崩热电子注入技术和高频Ｃ－Ｖ准静态Ｃ－Ｖ特性测试,研究了新型快速热氮化的ＳｉＯｘＮｙ介质膜界面陷阱的特征,侧重于研究界面陷阱的特性与分布。结果表明：这种ＳｉＯｘＮｙ薄膜禁带中央界面陷阱密度随氮化时间的分布变化呈现”回转效应“,且存在着不同类型、密度悬殊很大的电子陷阱、指出雪崩热电子注入过程中在Ｓｉ／ＳｉＯｘＮｙ界面上产生两种性质不同的快界面态陷阱;给出了这两种界面态陷阱密度在禁带中能量的分布     

电阻梯形网络电路的计算机辅助分析

在介绍电阻梯形网络电传统分析方法的基础上分析其不足,并根据阻元件的线性特征,提出了一种新的算法。     

运用科学方法 深化理科教改

通过对现代基础理论与实际问题应用关系的分析,提出了以科学方法论为主线,以开展数学建模教学为具体形式进行理论联系实际的教学改革设想。     

WAVELET PACKET AND FUZZY NEURAL NETWORK FOR TOOL CONDITION MONITORING *

在自动化制造中,线监控刀具状况以保护刀具与加工工件显得越来越重要．近年来,许多研究者在此领域进行了广泛的研究,然而由于加工过程的不确定性,现有的刀具监控系统的可靠性还有待提高．本文提出了一种基于小波包分析与模糊神经网络的自适应刀具监控系统,该系统利用小波包分析方法将加工过程振动信号分解为不同的频率段,并在此基础上,建立了自适应特征提取方法,为模糊神经网络提供最优的特征输入,然后模糊神经网络据此进行决策,分析刀具磨损状况．实验结果表明：该系统模糊神经网络能有效通过学习人类模糊知识和在线学习样本来提高刀具监控精度．     

粒状FeCu纳米颗粒体磁性和磁粘滞性的研究

溶胶法制备了不同粒度的ＦｅＣｕ纳米颗粒体,Ｘ射线衍射（ＸＲＤ）确定了样品的相组成和颗粒的平均粒度,振动样品磁强计（ＶＳＭ）测量了样品在不同温度下的磁化曲线,磁滞回线,涨落场和有效激活体积．“趋近饱和定律”方法求出了样品的有效磁各向异性常数．利用均匀反磁化模型计算了样品的矫顽力,计算值与测量值基本相符．结果表明,样品的激活体积随颗粒体积的增大而增大,激活体积与颗粒的平均体积的大小基本属于同一数量级,粒状ＦｅＣｕ纳米颗粒的反磁化基本符合均匀反磁化．     

A massive phytoplankton bloom induced by an ecosystem-scale iron fertilization experiment in the equatorial Pacific Ocean

The seeding of an expanse of surface waters in the equatorial Pacific Ocean with low concentrations of dissolved iron triggered a massive phytoplankton bloom which consumed large quantities of carbon dioxide and nitrate that these microscopic plants cannot fully utilize under natural conditions. These and other observations provide unequivocal support for the hypothesis that phytoplankton growth in this oceanic region is limited by iron bioavailability.     

Some citrus chitinases also possess chitosanase activities

Several acidic chitinase and chitosanase isoforms were found in 4-week-old nonembryogenic sweet orange (Valencia [Citrus sinensis (L.) Osbeck]) callus tissue. Two isoforms (designated A1-CF1 and A1-CF2) were purified to homogeneity using HPLC size exclusion, anion exchange, and chromatofocusing techniques. Both hydrolase isoforms exhibited activity with either colloidal chitin or solubilized shrimp shell chitosan. Specific activities for the purified isoforms could not be calculated because of the lack of protein and contamination of ampholytes. However, the specific activities for chitinase and chitosanase after anion exchange were respectively 404 nmol GlcNAc per min per mg protein and 2,475 nmol GlcN per min per mg protein. The M_r for both enzymes was 30,500. The homogeneous proteins cross-reacted in western blots with antiserum against a basic class I potato leaf chitinase.