Informatica

Multidimensional scaling (MDS) is a widely used technique for mapping data from a high-dimensional to a lower-dimensional space and for visualizing data. Recently, a new method, known as Geometric MDS, has been developed to minimize the MDS stress function by an iterative procedure, where coordinates of a particular point of the projected space are moved to the new position defined analytically. Such a change in position is easily interpreted geometrically. Moreover, the coordinates of points of the projected space may be recalculated simultaneously, i.e. in parallel, independently of each other. This paper has several objectives. Two implementations of Geometric MDS are suggested and analysed experimentally. The parallel implementation of Geometric MDS is developed for multithreaded multi-core processors. The sequential implementation is optimized for computational speed, enabling it to solve large data problems. It is compared with the SMACOF version of MDS. Python codes for both Geometric MDS and SMACOF are presented to highlight the differences between the two implementations. The comparison was carried out on several aspects: the comparative performance of Geometric MDS and SMACOF depending on the projection dimension, data size and computation time. Geometric MDS usually finds lower stress when the dimensionality of the projected space is smaller.

Multidimensional scaling with city-block distances is considered in this paper. The technique requires optimization of an objective function which has many local minima and can be non-differentiable at minimum points. This study is aimed at developing a fast and effective global optimization algorithm spanning the whole search domain and providing good solutions. A multimodal evolutionary algorithm is used for global optimization to prevent stagnation at bad local optima. Piecewise quadratic structure of the least squares objective function with city-block distances has been exploited for local improvement. The proposed algorithm has been compared with other algorithms described in literature. Through a comprehensive computational study, it is shown that the proposed algorithm provides the best results. The algorithm with fine-tuned parameters finds the global minimum with a high probability.

In this paper, the quality of quantization and visualization of vectors, obtained by vector quantization methods (self-organizing map and neural gas), is investigated. A multidimensional scaling is used for visualization of multidimensional vectors. The quality of quantization is measured by a quantization error. Two numerical measures for proximity preservation (Konig's topology preservation measure and Spearman's correlation coefficient) are applied to estimate the quality of visualization. Results of visualization (mapping images) are also presented.

Multidimensional scaling is a technique for exploratory analysis of multidimensional data widely usable in different applications. By means of this technique the image points in a low-dimensional embedding space can be found whose inter-point distances fit the given dissimilarities between the considered objects. In this paper dependence of relative visualization error on the dimensionality of embedding space is investigated. Both artificial and practical data sets have been used. The images in three-dimensional embedding space normally show the structural properties of sets of considered objects with acceptable accuracy, and widening of applications of stereo screens makes three-dimensional visualization very attractive.

We investigate applicability of quantitative methods to discover the most fundamental structural properties of the most reliable political data in Lithuania. Namely, we analyze voting data of the Lithuanian Parliament. Two most widely used techniques of structural data analysis (clustering and multidimensional scaling) are compared. We draw some technical conclusions which can serve as recommendations in more purposeful application of these methods.

In this paper, the relative multidimensional scaling method is investigated. This method is designated to visualize large multidimensional data. The method encompasses application of multidimensional scaling (MDS) to the so-called basic vector set and further mapping of the remaining vectors from the analyzed data set. In the original algorithm of relative MDS, the visualization process is divided into three steps: the set of basis vectors is constructed using the k-means clustering method; this set is projected onto the plane using the MDS algorithm; the set of remaining data is visualized using the relative mapping algorithm. We propose a modification, which differs from the original algorithm in the strategy of selecting the basis vectors. The experimental investigation has shown that the modification exceeds the original algorithm in the visualization quality and computational expenses. The conditions, where the relative MDS efficiency exceeds that of standard MDS, are estimated.

Recent publications on multidimensional scaling express contradicting opinion on multimodality of STRESS criterion. An example has been published with rigorously provable multimodality of STRESS. We present an example of data and the rigorous proof of multimodality of SSTRESS for this data. Some comments are included on widely accepted opinion that minimization of SSTRESS is easier than minimization of STRESS.