Interval estimation in discriminant analysis for large dimension

This paper is concerned with the interval estimation for the log odds of the posterior probability that the observation vector belongs to one of two homoscedastic multivariate normal distributions (Π₁ and Π₂). We give the limiting distribution of the unbiased estimator for the log odds as the sample sizes and the dimension jointly tend to infinity, and approximate the confidence interval based on the asymptotic distribution. Small-scale simulations are performed to check the precision of the approximation.     

飞行时间太赫兹断层成像

介绍了利用超短激光脉冲的高分辨飞行时间太赫兹断层成像的发展。其基本原理是,首先非线性光学晶体获得的宽带太赫兹脉冲进入到样品,随后检测每一层反射回来的回波脉冲。利用高斯窗口的去卷积过程我们可以获得清晰、高分辨的断层成像谱图。与聚四氟乙烯薄膜机械厚度测量的结果比较,如果一个样品的折射率为n,则轴向分辨率低于10/nμm。     

Factors affecting body size variation within a population of an herbivorous lady beetle,Henosepilachna niponica (Lewis)

Summary Size variation in newly-emerged adults was examined in two different local populations of an herbivorous lady beetle,Henosepilachna niponica, for 1976–80. Mean adult size of both sexes changed rather synchronously in the two populations over 5 years. Body size of adult beetles apparently decreased with increasing leaf damage of the plants on which they developed. Adult beetles which emerged late in the season, associated with increasing food deterioration, were smaller than those which emerged early. Ecological consequences of adult size variation is discussed in terms of oviposition site selection.     

Preference and performance linkage in aPhyllocolpa sawfly on the willow,Salix miyabeana, on Hokkaido

A leaf-folding sawfly in the genusPhyllocolpa (Hymenoptera: Tenthredinidae) attackingSalix miyabeana (Salicaceae) was studied near Sapporo, Hokkaido, along the Ishikari River in 1993. Host plant individuals were young trees 4–7 years old which were growing rapidly, producing some long shoots with large leaves. On a gradient of shoot length classes from 0–5 cm long to over 80 cm long, shoots were much more abundant in the shorter shoot length classes. However, attacks by ovipositing females increased as shoot length increase from 0 attacks on the shortest shoots to 5.17 attacks per shoot on the longest shoots. The frequency of attack per leaf increased from 0 to 0.13 over the same range of shoot lengths. This pattern of attack resulted in a high frequency of larval establishment in feeding sites, between 0.96 and 1.00, in all attacked shoot length categories. However, probability of survival to a late instar larva increased with shoot length and corresponded to the attack pattern, indicating a preference-performance linkage between female ovipositional decisions and larval survival. The patterns found for thisPhyllocolpa species are similar for galling sawflies in North America and Europe, especially in the genusEuura, members of which make stem, bud and leaf midrib galls. Extending the pattern to aPhyllocolpa species broadens identification of pattern and ultimately the generality of the emerging theory on populations of galling sawflies.     

How to Build an Intentional Android: Infants' Imitation of a Robot's Goal‐Directed Actions

This study examined whether young children are able to imitate a robot's goal‐directed actions. Children (24–35 months old) viewed videos showing a robot attempting to manipulate an object (e.g., putting beads inside a cup) but failing to achieve its goal (e.g., beads fell outside the cup). In 1 video, the robot made eye contact with a human before and after it failed the action. In another video, the robot did not make eye contact with the human adult. Only in the former condition did children “imitate” the robot's “intended” but unconsummated actions (e.g., putting beads inside a cup). When the robot did not make eye contact, children performed poorly, at the baseline level. These results suggest that human‐like gaze behaviors, not human‐like morphology, may play an important role in young children's imitation of a nonhuman agent's goal‐directed behaviors.     

The Asymptotic Approximation of EPMC for Linear Discriminant Rules Using a Moore-Penrose Inverse Matrix in High Dimension

We consider the discriminant rule in a high-dimensional setting, i.e., when the number of feature variables p is comparable to or larger than the number of observations N. The discriminant rule must be modified in order to cope with singular sample covariance matrix in high-dimension. One way to do so is by considering the Moor-Penrose inverse matrix. Recently, Srivastava (2006 Srivastava , M. S. ( 2006 ). Minimum distance classification rules for high dimensional data . J. Multivariate Anal. 97 : 2057 – 2070 .[Crossref], [Web of Science ®] , [Google Scholar]) proposed maximum likelihood ratio rule by using Moor-Penrose inverse matrix of sample covariance matrix. In this article, we consider the linear discriminant rule by using Moor-Penrose inverse matrix of sample covariance matrix (LDRMP). With the discriminant rule, the expected probability of misclassification (EPMC) is commonly used as measure of the classification accuracy. We investigate properties of EPMC for LDRMP in high-dimension as well as the one of the maximum likelihood rule given by Srivastava (2006 Srivastava , M. S. ( 2006 ). Minimum distance classification rules for high dimensional data . J. Multivariate Anal. 97 : 2057 – 2070 .[Crossref], [Web of Science ®] , [Google Scholar]). From our asymptotic results, we show that the classification accuracy of LDRMP depends on new distance. Additionally, our asymptotic result is verified by using the Monte Carlo simulation.     

An experimental analysis of the relationship between species combination and community persistence

Among many stabilizing factors for community dynamics, nonlinear biological interactions such as type III functional response have been widely considered to be major characteristics. However, most experimental biological communities employed so far had quite simple structures. Therefore, the possibility that the conclusions in earlier studies were dependent on simple community structure is undeniable. In this study, using a multiple-species experimental community, we evaluated which combinations of component species and what kinds of interspecific interactions allow communities to persist and how these contribute to community persistence. We conducted experimental communities using two species of beans, the adzuki bean (Vigna angularis) and the red kidney bean (Phaseolus vulgaris), two species of bean weevils, the Mexican bean weevil (Zabrotes subfasciatus, Coleoptera: Bruchidae) and adzuki bean weevil (Callosobruchus chinensis, Coleoptera: Bruchidae), and two species of parasitic wasp, Heterospilus prosopidis (Hymenoptera: Braconidae) and Anisopteromalus calandrae (Hymenoptera: Pteromalidae). The outcome of multiple-generation experimental communities was explained by the characteristics of component species obtained from short-term experiments. In our two resources–two herbivores–one carnivore system, the strong density-dependent attack ability of one parasitic wasp species (A. calandrae) led to the extinction of C. chinensis. On the other hand, the weak density-dependent attack ability of the other parasitic wasp species (H. prosopidis) led to system persistence. Our overall results show that, in a multiple-species community, the combination of species itself is more important for community persistence than are the characteristics of the particular species. Received: September 29, 1997 / Accepted: October 5, 1998