A novel kernel Fisher discriminant analysis: constructing informative kernel by decision tree ensemble for metabolomics data analysis

Large amounts of data from high-throughput metabolomics experiments become commonly more and more complex, which brings an enormous amount of challenges to existing statistical modeling. Thus there is a need to develop statistically efficient approach for mining the underlying metabolite information contained by metabolomics data under investigation. In the work, we developed a novel kernel Fisher discriminant analysis (KFDA) algorithm by constructing an informative kernel based on decision tree ensemble. The constructed kernel can effectively encode the similarities of metabolomics samples between informative metabolites/biomarkers in specific parts of the measurement space. Simultaneously, informative metabolites or potential biomarkers can be successfully discovered by variable importance ranking in the process of building kernel. Moreover, KFDA can also deal with nonlinear relationship in the metabolomics data by such a kernel to some extent. Finally, two real metabolomics datasets together with a simulated data were used to demonstrate the performance of the proposed approach through the comparison of different approaches.     

高光谱数据的刚竹毒蛾虫害程度检测

刚竹毒蛾虫害检测对毛竹的生长和竹业的发展起着至关重要的作用。根据高光谱冠层光谱信息与刚竹毒蛾虫害程度之间的关系,提取冠层光谱中与虫害紧密相关的特征波长、指数以及光谱参数等,利用Fisher判别分析法建立刚竹毒蛾虫害程度检测模型。分别以原始光谱的400～508,586～693和724～900 nm处的波长、包络线去除光谱的400～756 nm之间的特征波长、9种冠层光谱植被指数和7种冠层特征光谱参数作为Fisher判别函数自变量,构建判别函数。收集300组毛竹叶片虫害样本数据,随机划分为210组建模集与90组验证集,根据检测精度、Kappa系数以及判定系数R2作为检验标准,对建立的判别函数进行效果评价与对比。结果表明,以原始光谱、去包络线光谱、冠层指数、光谱参数为自变量建立的Fisher判别函数的检验精度分别为：84.4%,81.1%,79.7%,78.7%;Kappa系数分别为：0.79,0.74,0.74,0.76;R2分别为：0.89,0.88,0.88和0.85。由此可知,Fisher判别分析模型建立的函数具备很好的刚竹毒蛾虫害程度检测能力,而且基于冠层原始光谱建立的判别函数检测效果最佳。根据高光谱数据的冠层原始光谱建立的判别函数对福建省顺昌县大干镇武坊村的洋门和土垅村的上湖竹林进行刚竹毒蛾虫害程度检测。检测结果为：上湖两个样区的竹林以健康为主。洋门两个样区虫害程度以中度和重度为主。因此基于无人机高光谱遥感对于刚竹毒蛾虫害的大面积检测具有可行性,该方法可为虫害检测的探究提供参考,为基于冠层遥感虫害检测贡献理论支撑。     

Information-theoretical complexity for the hydrogenic abstraction reaction

In this work, we have investigated the complexity of the hydrogenic abstraction reaction by means of information functionals such as disequilibrium (D), exponential entropy (L), Fisher information (I), power entropy (J) and joint information-theoretic measures, i.e. the I–D, D–L and I–J planes and the Fisher–Shannon and López–Mancini–Calbet (LMC) shape complexities. The analysis of the information-theoretical functionals of the one-particle density was computed in position (r) and momentum (p) space. The analysis revealed that all of the chemically significant regions can be identified from the information functionals and most of the information-theoretical planes, i.e. the reactant/product regions (R/P), the transition state (TS), including those that are not present in the energy profile such as the bond cleavage energy region (BCER), and the bond breaking/forming regions (B–B/F). The analysis of the complexities shows that, in position as well as in the joint space, the energy profile of the abstraction reaction bears the same information-theoretical features as the LMC and FS measures. We discuss why most of the chemical features of interest, namely the BCER and B–B/F, are lost in the energy profile and that they are only revealed when particular information-theoretical aspects of localizability (L or J), uniformity (D) and disorder (I) are considered.     

超衍射成像中 双螺旋点扩展函数的三维定位精度

开关效应和单分子定位的结合可以实现样品的超衍射分辨成像, 双螺旋点扩展函数将单分子定位纳米分辨从二维扩展到了三维.本文对双螺旋点扩展函数的三维定位精度展开了探讨.首先, 基于费希尔信息量, 计算了双螺旋点扩展函数的无偏估计, 得出其理论定位精度, 并分析了光子数、背景噪声以及有效像元尺寸大小对其定位精度的影响; 其次, 基于单分子定位实验过程中对于数据分析通常采用的高斯拟合质心定位算法, 通过误差传递函数定律求得双螺旋点扩展函数的轴向定位精度.计算机模拟结果表明, 在光子数大于1000的条件下, 高斯拟合质心定位精度和费希尔信息量理论定位精度符合较好. 本文的讨论不仅为双螺旋点扩展函数的三维定位精度提供了理论依据, 同时也可为实验提供理论指导. 关键词： 双螺旋点扩展函数 费希尔信息量 定位精度 高斯拟合     

Uncertainty principle and quantum Fisher information

A family of inequalities, related to the uncertainty principle, has been recently proved by S. Luo, Z. Zhang, Q. Zhang, H. Kosaki, K. Yanagi, S. Furuichi and K. Kuriyama. We show that the inequalities have a geometric interpretation in terms of quantum Fisher information. Using this formulation one may naturally ask if this family of inequalities can be further extendend, for example to the RLD quantum Fisher information. We show that this is impossible by producing a family of counterexamples.     

左截断数据中的Fisher信息量

本文研究左截断数据(U,V)含θ的Fisher信息量,并给出了基于分布和失效率表示的两个表达式.特别在Koziol-Green模型下研究左截断观测数据(U,V)含θ的Fisher信息量与完全数据X含θ的Fisher信息量的关系,得出在该模型下左截断观测数据(U,V)含θ的Fisher信息量不仅未损失,反而增加的结论.     

Global statistical information in exponential experiments and selection of exponential models

The concept of global statistical information in the classical statistical experiment with independent exponentially distributed samples is investigated. Explicit formulas are evaluated for common exponential families. It is shown that the generalized likelihood ratio test procedure of model selection can be replaced by a generalized information procedure. Simulations in a classical regression model are used to compare this procedure with that based on the Akaike criterion.     

Direct approach to quantum extensions of Fisher information

By manipulating classical Fisher information and employing various derivatives of density operators, and using entirely intuitive and direct methods, we introduce two families of quantum extensions of Fisher information that include those defined via the symmetric logarithmic derivative, via the right logarithmic derivative, via the Bogoliubov-Kubo-Mori derivative, as well as via the derivative in terms of commutators, as special cases. Some fundamental properties of these quantum extensions of Fisher information are investigated, a multi-parameter quantum Cramér-Rao inequality is established, and applications to characterizing quantum uncertainty are illustrated.      

Inference for earthquake models: A self-correcting model

Questions of asymptotic inference are discussed for a point process model in which the conditional intensity function increases monotonically between events and drops by determined (nonrandom) amounts after each event. Parameter estimates are shown to be consistent and, except under the null hypothesis of a Poisson process, normally distributed. Under the null hypothesis, however, the Hessian matrix is not asymptotically constant, and the limiting distribution of the likelihood ratio statistics is not χ², but has a form related to that of the Cramer-von Mises ω² statistic for the test of goodness of fit.     

Limiting properties of some measures of information

In this paper we investigate the limiting behaviour of the measures of information due to Csiszár, Rényi and Fisher. Conditions for convergence of measures of information and for convergence of Radon-Nikodym derivatives are obtained. Our results extend the results of Kullback (1959,Information Theory and Statistics, Wiley, New York) and Kirmani (1971,Ann. Inst. Statist. Math.,23, 157–162).