We give a tutorial overview of several geometric methods for dimension reduction. We divide the methods into projective methods and methods that model the manifold on which the data lies. For projective methods, we review projection pursuit, principal component analysis (PCA), kernel PCA, probabilistic PCA, canonical correlation analysis, oriented PCA, and several techniques for sufficient dime...

Y chromosomal STRs show sufficient variability among individduals in a population and a high degree of geographical differentiation, such that their polymorphic character makes them especially suited for population genetic studies. To investígate the polymorphism of a set of 17 Y-STR loci in northern China, we genotyped the 17 Y chromosomal STR loci in a population sample of 377 unrelated males...

Localization, an important challenge in wireless sensor networks, is the process of sensor nodes self-determining their position. The difficulty encountered is in cost-effectively providing acceptable accuracy in localization. The potential for the deployment of high density networks in the near future makes scalability a critical issue in localization. In this paper we propose Cluster-based Lo...

| Multidimensional Scaling (MDS) techniques always pose the problem of analysing a large number N of points, without collecting all N(N?1) 2 possible interstimuli dissimilarities, and while keeping satisfactory solutions. In the case of metric MDS, it was found that a theoretical minimum of appropriate 2N ?3 exact Euclidean distances are suf-cient for the unique representation of N points in a ...

Distance-based methods in machine learning and pattern recognition have to rely on a metric distance between points in the input space. Instead of specifying a metric a priori, we seek to learn the metric from data via kernel methods and multidimensional scaling (MDS) techniques. Under the classification setting, we define discriminant kernels on the joint space of input and output spaces and p...

In this study we show that humans are able to form a perceptual space from a complex, three-dimensional shape space that is highly congruent to the physical object space no matter if the participants explore the objects visually or haptically. The physical object space consists of complex, shell-shaped objects which were generated by varying three shape parameters. In several psychophysical exp...

Multidimensional scaling (MDS) is a data analysis technique for representing measurements of (dis)similarity among pairs of objects as distances between points in a low-dimensional space. MDS methods differ mainly according to the distance model used to scale the proximities. The most usual model is the Euclidean one, although a spherical model is often preferred to represent correlation measur...

This article is an updated version of De Leeuw and Mair (2009b) published in the Journal of Statistical Software. It elaborates on the methodology of multidimensional scaling problems (MDS) solved by means of the majorization algorithm. The objective function to be minimized is known as stress and functions which majorize stress are elaborated. This strategy to solve MDS problems is called SMAC...

This report discusses one paper for linear data dimensionality reduction, Eigenfaces, and two recently developed nonlinear techniques. The first nonlinear method, Locally Linear Embedding (LLE), maps the input data points to a single global coordinate system of lower dimension in a manner that preserves the relationships between neighboring points. The second method, Isomap, computes geodesic d...

Two experiments tested the hypothesis that implicit personality theory person types are composed of causally interconnected traits. Experiment 1 showed that the weakest trait member of a person type is perceived as more causally related to the core trait members of the type than are nonmember traits, even when those nonmember traits are both more highly correlated with and closer in multidimens...