Discriminative Orthogonal Neighborhood-Preserving Projections for Classification

Orthogonal neighborhood-preserving projection (ONPP) is a recently developed orthogonal linear algorithm for overcoming the out-of-sample problem existing in the well-known manifold learning algorithm, i.e., locally linear embedding. It has been shown that ONPP is a strong analyzer of high-dimensional data. However, when applied to classification problems in a supervised setting, ONPP only focuses on the intraclass geometrical information while ignores the interaction of samples from different classes. To enhance the performance of ONPP in classification, a new algorithm termed discriminative ONPP (DONPP) is proposed in this paper. DONPP 1) takes into account both intraclass and interclass geometries; 2) considers the neighborhood information of interclass relationships; and 3) follows the orthogonality property of ONPP. Furthermore, DONPP is extended to the semisupervised case, i.e., semisupervised DONPP (SDONPP). This uses unlabeled samples to improve the classification accuracy of the original DONPP. Empirical studies demonstrate the effectiveness of both DONPP and SDONPP.

Selection for variables in quantitative structure-activity/property relationship studies - Comparison between methods of orthogonal descriptors and leaps-and-bounds regression analysis

In this paper, the comparison of orthogonal descriptors and Leaps-and-Bounds regression analysis is performed. The results obtained by using orthogonal descriptors are better than that obtained by using Leaps-and-Bounds regression for the data set of nitrobenzenes used in this study. Leaps-and-Bounds regression can be used effectively for selection of variables in quantitative structure-activity/property relationship(QSAR/QSPR) studies. Consequently, orthogonalisation of descriptors is also a good method for variable selection for studies on QSAR/QSPR.

Variable selection by orthogonal descriptors and prediction of R-f value of phenol and aniline derivatives

Orthogonal descriptors is a viable method for variable selection, but this method strongly depend on the orthogonalisation ordering of the descriptors. In this paper, we compared the different methods used for order the descriptors. It showed that better results could be achieved with the use of backward elimination ordering. We predicted R-f value of phenol and aniline derivatives by this method, and compared it with classical algorithms such as forward selection, backward elimination, and stepwise procedure. Some interesting hints were obtained.

Studies on the hemolytic activity of tentacle extracts of jellyfish Rhopilema esculentum Kishinouye: Application of orthogonal test

The present work is first reporting the hemolytic activity of venom from jellyfish Rhopilema esculentum Kishinouye extracted by different phosphate buffer solutions and incubated at different temperature according to the orthogonal test L6(1) x 3(6). Of the seven controllable independent variables, incubated temperature and phenylmethylsulfonyl fluoride (PMSF) had strongest effect on the hemolytic activity. (c) 2006 Elsevier B.V. All rights reserved.

Orthogonal orientation control of carbon nanotube growth.

Carbon nanotubes (CNTs) have attracted attention for their remarkable electrical properties and have being explored as one of the best building blocks in nano-electronics. A key challenge to realize such potential is the control of the nanotube growth directions. Even though both vertical growth and controlled horizontal growth of carbon nanotubes have been realized before, the growth of complex nanotube structures with both vertical and horizontal orientation control on the same substrate has never been achieved. Here, we report a method to grow three-dimensional (3D) complex nanotube structures made of vertical nanotube forests and horizontal nanotube arrays on a single substrate and from the same catalyst pattern by an orthogonally directed nanotube growth method using chemical vapor deposition (CVD). More importantly, such a capability represents a major advance in controlled growth of carbon nanotubes. It enables researchers to control the growth directions of nanotubes by simply changing the reaction conditions. The high degree of control represented in these experiments will surely make the fabrication of complex nanotube devices a possibility.

Reformulating time-dependent density functional theory with non-orthogonal localized molecular orbitals.

Time-dependent density functional theory (TDDFT) has broad application in the study of electronic response, excitation and transport. To extend such application to large and complex systems, we develop a reformulation of TDDFT equations in terms of non-orthogonal localized molecular orbitals (NOLMOs). NOLMO is the most localized representation of electronic degrees of freedom and has been used in ground state calculations. In atomic orbital (AO) representation, the sparsity of NOLMO is transferred to the coefficient matrix of molecular orbitals (MOs). Its novel use in TDDFT here leads to a very simple form of time propagation equations which can be solved with linear-scaling effort. We have tested the method for several long-chain saturated and conjugated molecular systems within the self-consistent charge density-functional tight-binding method (SCC-DFTB) and demonstrated its accuracy. This opens up pathways for TDDFT applications to large bio- and nano-systems.

Bivariate distribution modeling with tree diameter and height data

The Logit-Logistic (LL), Johnson's SB, and the Beta (GBD) are flexible four-parameter probability distribution models in terms of the (skewness-kurtosis) region covered, and each has been used for modeling tree diameter distributions in forest stands. This article compares bivariate forms of these models in terms of their adequacy in representing empirical diameter-height distributions from 102 sample plots. Four bivariate models are compared: SBB, the natural, well-known, and much-used bivariate generalization of SB; the bivariate distributions with LL, SB, and Beta as marginals, constructed using Plackett's method (LL-2P, etc.). All models are fitted using maximum likelihood, and their goodness-of-fits are compared using minus log-likelihood (equivalent to Akaike's Information Criterion, the AIC). The performance ranking in this case study was SBB, LL-2P, GBD-2P, and SB-2P