An interview with Sam C. Saunders

Sam C. Saunders, the son of Elizabeth Cundiff and Winston E. Saunders, was born in Richland, OR, on February 24, 1931. The family moved to La Grande, OR, in 1944, where Sam completed high school and two years at Eastern Oregon College. He then received the BSc degree in Mathematics from the University of Oregon, Eugene, OR, in 1952, and he attended the University of Washington, Seattle, WA, receiving a PhD degree under Z. W. Birnbaum. After graduating, he accepted employment at the Boeing Company in its Mathematical Services Unit and, in 1972, a position as a Full Professor at Washington State University, Pullman, WA, from which he retired in 1996.     

Synergistic Effects of Pseudocolor Imaging,Differentiation, and Square and Logarithmic Conversion on Accuracy of Quantification of Chemical Characteristics Using Test Strips and Similar Products

Test strips and similar products are highly feasible tools for the rapid and approximate determination of chemical characteristics. Although the application of both the quantitative observation of coloration and regression modeling has recently enabled these products to become quantitative tools, their precision and accuracy may be further improved. In this study, the pseudocolor imaging of the coloration image, derivative spectrophotometry-like differentiation of the coloration values, and logarithmic conversion of the raw and derivative values were compared in terms of the precision and accuracy of the quantitative determination of corrosiveness, glucose, nitrate, and pH using the products. The best regression models for the determination were provided by the combination of pseudocolor imaging and differentiation (nitrate and pH); pseudocolor imaging, differentiation, and square-conversion (corrosiveness); or all of the techniques (glucose). When compared to the use of the original 10 raw coloration variables of red-green-blue, cyan-magenta-yellow-key black, and L*a*b* color models only, the above combinations improved the normalized mean absolute error from 14.8% to 3.09% (corrosiveness), 6.33% to 3.15% (glucose), 7.46% to 4.56% (nitrate), and 3.22% to 0.94% (pH). These achievements were largely attributed to the combination of multiple variables that have non-linear and nonmonotonic relationships with the chemical characteristics.     

Statistics of reaction progress variable and mixture fraction gradients of a pulverised coal jet flame using Direct Numerical Simulation data

The statistical behaviour of the variances, covariance and gradients of the reaction progress variable (c), and the mixture fraction (ξ) have been analysed in a pulverised coal jet flame using a three-dimensional carrier phase direct numerical simulation (DNS) dataset. It has been observed that the Favre-probability density functions (PDFs) of c and ξ can be parametrised by the standard β function. Furthermore, the log-normal distribution has been found to accurately represent |?c| and |?ξ|. It is also found that ?c and ?ξ remain aligned throughout the flame brush. Finally the joint PDF of |?c| and |?ξ| has been compared with the product of the PDF of |?c| and PDF of |?ξ| extracted from carrier phase DNS, and it has been found that |?c| and |?ξ| are not statistically independent in the case investigated.The bivariate log-normal distributions with and without correlation have also been considered, and the former is found to be in better agreement with the carrier phase DNS data.     

Self-affinity and nonextensivity of sunspots

In this paper we study the time series of sunspots by using two different approaches, analyzing its self-affine behavior and studying its distribution. The long-range correlation exponent α has been calculated via Detrended Fluctuation Analysis and the power law vanishes to values greater than 11 years. On the other hand, the distribution of the sunspots obeys a q-exponential decay that suggests a non-extensive behavior. This observed characteristic seems to take an alternative interpretation of the sunspots dynamics. The present findings suggest us to propose a dynamic model of sunspots formation based on a nonlinear Fokker–Planck equation. Therefore its dynamic process follows the generalized thermostatistical formalism.     

A new Monte Carlo model for predicting the mechanical properties of fiber yarns

Understanding the complicated failure mechanisms of hierarchical composites such as fiber yarns is essential for advanced materials design. In this study, we developed a new Monte Carlo model for predicting the mechanical properties of fiber yarns that includes statistical variation in fiber strength. Furthermore, a statistical shear load transfer law based on the shear lag analysis was derived and implemented to simulate the interactions between adjacent fibers and provide a more accurate tensile stress distribution along the overlap distance. Simulations on two types of yarns, made from different raw materials and based on distinct processing approaches, predict yarn strength values that compare favorably with experimental measurements. Furthermore, the model identified very distinct dominant failure mechanisms for the two materials, providing important insights into design features that can improve yarn strength.     

Differentiation between Shallow and Deep Charge Trap States on Single Poly(3‐hexylthiophene) Chains through Fluorescence Photon Statistics

Blinking of the photoluminescence (PL) emitted from individual conjugated polymer chains is one of the central observations made by single‐molecule spectroscopy (SMS). Important information, for example regarding excitation energy transfer, can be extracted by evaluating dynamic quenching. However, the nature of trap states, which are responsible for PL quenching, often remains obscured. We present a detailed investigation of the photon statistics of single poly(3‐hexylthiophene) (P3HT) chains obtained by SMS. The photon statistics provide a measure of the number and brightness of independently emitting areas on a single chain. These observables can be followed during blinking. A decrease in PL intensity is shown to be correlated with either 1) a decrease in the average brightness of the emitting sites; or 2) a decrease in the number of emitting regions. We attribute these phenomena to the formation of 1) shallow charge traps, which can weakly affect all emitting areas of a single chain at once; and 2) deep traps, which have a strong effect on small regions within the single chains.     

Odd length for even hyperoctahedral groups and signed generating functions

We define a new statistic on the even hyperoctahedral groups which is a natural analogue of the odd length statistic recently defined and studied on Coxeter groups of types $A$ and $B$. We compute the signed (by length) generating function of this statistic over the whole group and over its maximal and some other quotients and show that it always factors nicely. We also present some conjectures.     

Logistic回归模型的统计诊断

统计诊断的主要任务就是通过诊断统计量检测已知观测数据在用既定模型拟合时的合理性,主要是找出数据当中的异常点或强影响点。本文主要研究Logostic回归模型的诊断统计量和诊断统计图。用牛顿迭代法给出Logistic回归模型的极大似然估计值,根据扰动模型得到传统的诊断统计量,结合残差、杠杆值和系数变化三者构造新的诊断统计量,绘制新的诊断统计图,通过模拟研究说明新的诊断统计量的有效性,最后用一个实际案例说明新的诊断方法的应用并进一步验证其优越性。     

One-Step Estimator Paths for Concave Regularization

The statistics literature of the past 15 years has established many favorable properties for sparse diminishing-bias regularization: techniques that can roughly be understood as providing estimation under penalty functions spanning the range of concavity between ?₀ and ?₁ norms. However, lasso ?₁-regularized estimation remains the standard tool for industrial Big Data applications because of its minimal computational cost and the presence of easy-to-apply rules for penalty selection. In response, this article proposes a simple new algorithm framework that requires no more computation than a lasso path: the path of one-step estimators (POSE) does ?₁ penalized regression estimation on a grid of decreasing penalties, but adapts coefficient-specific weights to decrease as a function of the coefficient estimated in the previous path step. This provides sparse diminishing-bias regularization at no extra cost over the fastest lasso algorithms. Moreover, our gamma lasso implementation of POSE is accompanied by a reliable heuristic for the fit degrees of freedom, so that standard information criteria can be applied in penalty selection. We also provide novel results on the distance between weighted-?₁ and ?₀ penalized predictors; this allows us to build intuition about POSE and other diminishing-bias regularization schemes. The methods and results are illustrated in extensive simulations and in application of logistic regression to evaluating the performance of hockey players. Supplementary materials for this article are available online.