Weighted metric multidimensional scaling

This paper establishes a general framework for metric scaling of any distance measure between individuals based on a rectangular individuals-by-variables data matrix. The method allows visualization of both individuals and variables as well as preserving all the good properties of principal axis methods such as principal components and correspondence analysis, based on the singular-value decomposition, including the decomposition of variance into components along principal axes which provide the numerical diagnostics known as contributions. The idea is inspired from the chi-square distance in correspondence analysis which weights each coordinate by an amount calculated from the margins of the data table. In weighted metric multidimensional scaling (WMDS) we allow these weights to be unknown parameters which are estimated from the data to maximize the fit to the original distances. Once this extra weight-estimation step is accomplished, the procedure follows the classical path in decomposing a matrix and displaying its rows and columns in biplots.

Nonlinear principal and canonical directions from continuous extensions of multidimensional scaling

A continuous random variable is expanded as a sum of a sequence of uncorrelated random variables. These variables are principal dimensions in continuous scaling on a distance function, as an extension of classic scaling on a distance matrix. For a particular distance, these dimensions are principal components. Then some properties are studied and an inequality is obtained. Diagonal expansions are considered from the same continuous scaling point of view, by means of the chi-square distance. The geometric dimension of a bivariate distribution is defined and illustrated with copulas. It is shown that the dimension can have the power of continuum.

Comparing methods for dimensionality reduction when data are density functions

Functional Data Analysis (FDA) deals with samples where a whole function is observedfor each individual. A particular case of FDA is when the observed functions are densityfunctions, that are also an example of infinite dimensional compositional data. In thiswork we compare several methods for dimensionality reduction for this particular typeof data: functional principal components analysis (PCA) with or without a previousdata transformation and multidimensional scaling (MDS) for diferent inter-densitiesdistances, one of them taking into account the compositional nature of density functions. The difeerent methods are applied to both artificial and real data (householdsincome distributions)

Analyzing shapes as compositions of distances

We propose to analyze shapes as “compositions” of distances in Aitchison geometry asan alternate and complementary tool to classical shape analysis, especially when sizeis non-informative.Shapes are typically described by the location of user-chosen landmarks. Howeverthe shape – considered as invariant under scaling, translation, mirroring and rotation– does not uniquely define the location of landmarks. A simple approach is to usedistances of landmarks instead of the locations of landmarks them self. Distances arepositive numbers defined up to joint scaling, a mathematical structure quite similar tocompositions. The shape fixes only ratios of distances. Perturbations correspond torelative changes of the size of subshapes and of aspect ratios. The power transformincreases the expression of the shape by increasing distance ratios. In analogy to thesubcompositional consistency, results should not depend too much on the choice ofdistances, because different subsets of the pairwise distances of landmarks uniquelydefine the shape.Various compositional analysis tools can be applied to sets of distances directly or afterminor modifications concerning the singularity of the covariance matrix and yield resultswith direct interpretations in terms of shape changes. The remaining problem isthat not all sets of distances correspond to a valid shape. Nevertheless interpolated orpredicted shapes can be backtransformated by multidimensional scaling (when all pairwisedistances are used) or free geodetic adjustment (when sufficiently many distancesare used)

Unfolding a symmetric matrix

Graphical displays which show inter--sample distances are importantfor the interpretation and presentation of multivariate data. Except whenthe displays are two--dimensional, however, they are often difficult tovisualize as a whole. A device, based on multidimensional unfolding, isdescribed for presenting some intrinsically high--dimensional displays infewer, usually two, dimensions. This goal is achieved by representing eachsample by a pair of points, say $R_i$ and $r_i$, so that a theoreticaldistance between the $i$-th and $j$-th samples is represented twice, onceby the distance between $R_i$ and $r_j$ and once by the distance between$R_j$ and $r_i$. Self--distances between $R_i$ and $r_i$ need not be zero.The mathematical conditions for unfolding to exhibit symmetry are established.Algorithms for finding approximate fits, not constrained to be symmetric,are discussed and some examples are given.

Dynamic perceptual mapping

Perceptual maps have been used for decades by market researchers to illuminatethem about the similarity between brands in terms of a set of attributes, to position consumersrelative to brands in terms of their preferences, or to study how demographic and psychometricvariables relate to consumer choice. Invariably these maps are two-dimensional and static. Aswe enter the era of electronic publishing, the possibilities for dynamic graphics are opening up.We demonstrate the usefulness of introducing motion into perceptual maps through fourexamples. The first example shows how a perceptual map can be viewed in three dimensions,and the second one moves between two analyses of the data that were collected according todifferent protocols. In a third example we move from the best view of the data at the individuallevel to one which focuses on between-group differences in aggregated data. A final exampleconsiders the case when several demographic variables or market segments are available foreach respondent, showing an animation with increasingly detailed demographic comparisons.These examples of dynamic maps use several data sets from marketing and social scienceresearch.

Measuring subcompositional incoherence

Subcompositional coherence is a fundamental property of Aitchison s approach to compositional data analysis, and is the principal justification for using ratios of components. We maintain, however, that lack of subcompositional coherence, that is incoherence, can be measured in an attempt to evaluate whether any given technique is close enough, for all practical purposes, to being subcompositionally coherent. This opens up the field to alternative methods, which might be better suited to cope with problems such as data zeros and outliers, while being only slightly incoherent. The measure that we propose is based on the distance measure between components. We show that the two-part subcompositions, which appear to be the most sensitive to subcompositional incoherence, can be used to establish a distance matrix which can be directly compared with the pairwise distances in the full composition. The closeness of these two matrices can be quantified using a stress measure that is common in multidimensional scaling, providing a measure of subcompositional incoherence. The approach is illustrated using power-transformed correspondence analysis, which has already been shown to converge to log-ratio analysis as the power transform tends to zero.

Weighted Euclidean biplots

We construct a weighted Euclidean distance that approximates any distance or dissimilarity measure between individuals that is based on a rectangular cases-by-variables data matrix. In contrast to regular multidimensional scaling methods for dissimilarity data, the method leads to biplots of individuals and variables while preserving all the good properties of dimension-reduction methods that are based on the singular-value decomposition. The main benefits are the decomposition of variance into components along principal axes, which provide the numerical diagnostics known as contributions, and the estimation of nonnegative weights for each variable. The idea is inspired by the distance functions used in correspondence analysis and in principal component analysis of standardized data, where the normalizations inherent in the distances can be considered as differential weighting of the variables. In weighted Euclidean biplots we allow these weights to be unknown parameters, which are estimated from the data to maximize the fit to the chosen distances or dissimilarities. These weights are estimated using a majorization algorithm. Once this extra weight-estimation step is accomplished, the procedure follows the classical path in decomposing the matrix and displaying its rows and columns in biplots.

A comparison of rapid bioassessment protocols used in 2 regions with Mediterranean climates, the Iberian Peninsula and South Africa

Abstract. The ability of 2 Rapid Bioassessment Protocols (RBPs) to assess stream water quality was compared in 2 Mediterranean-climate regions. The most commonly used RBPs in South Africa (SAprotocol) and the Iberian Peninsula (IB-protocol) are both multihabitat, field-based methods that use macroinvertebrates. Both methods use preassigned sensitivity weightings to calculate metrics and biotic indices. The SA- and IB-protocols differ with respect to sampling equipment (mesh size: 1000 lm vs 250 300 lm, respectively), segregation of habitats (substrate vs flow-type), and sampling and sorting procedures (variable time and intensity). Sampling was undertaken at 6 sites in South Africa and 5 sites in the Iberian Peninsula. Forty-four and 51 macroinvertebrate families were recorded in South Africa and the Iberian Peninsula, respectively; 77.3% of South African families and 74.5% of Iberian Peninsula families were found using both protocols. Estimates of community similarity compared between the 2 protocols were .60% similar among sites in South Africa and .54% similar among sites in the Iberian Peninsula (BrayCurtis similarity), and no significant differences were found between protocols (Multiresponse Permutation Procedure). Ordination based on Non-metric Multidimensional Scaling grouped macroinvertebrate samples on the basis of site rather than protocol. Biotic indices generated with the 2 protocols at each site did not differ. Thus, both RBPs produced equivalent results, and both were able to distinguish between biotic communities (mountain streams vs foothills) and detect water-quality impairment, regardless of differences in sampling equipment, segregation of habitats, and sampling and sorting procedures. Our results indicate that sampling a single habitat may be sufficient for assessing water quality, but a multihabitat approach to sampling is recommended where intrinsic variability of macroinvertebrate assemblages is high (e.g., in undisturbed sites in regions with Mediterranean climates). The RBP of choice should depend on whether the objective is routine biomonitoring of water quality or autecological or faunistic studies.

Searching for consensus in ahp-group decision making. A bayesian perspective

This paper sets out to identify the initial positions of the different decisionmakers who intervene in a group decision making process with a reducednumber of actors, and to establish possible consensus paths between theseactors. As a methodological support, it employs one of the most widely-knownmulticriteria decision techniques, namely, the Analytic Hierarchy Process(AHP). Assuming that the judgements elicited by the decision makers follow theso-called multiplicative model (Crawford and Williams, 1985; Altuzarra et al.,1997; Laininen and Hämäläinen, 2003) with log-normal errors and unknownvariance, a Bayesian approach is used in the estimation of the relative prioritiesof the alternatives being compared. These priorities, estimated by way of themedian of the posterior distribution and normalised in a distributive manner(priorities add up to one), are a clear example of compositional data that will beused in the search for consensus between the actors involved in the resolution ofthe problem through the use of Multidimensional Scaling tools

Group preference structures in AHP–group decision making

This paper presents a procedure that allows us to determine the preference structures(PS) associated to each of the different groups of actors that can be identified in a groupdecision making problem with a large number of individuals. To that end, it makesuse of the Analytic Hierarchy Process (AHP) (Saaty, 1980) as the technique to solvediscrete multicriteria decision making problems. This technique permits the resolutionof multicriteria, multienvironment and multiactor problems in which subjective aspectsand uncertainty have been incorporated into the model, constructing ratio scales correspondingto the priorities relative to the elements being compared, normalised in adistributive manner (wi = 1). On the basis of the individuals’ priorities we identifydifferent clusters for the decision makers and, for each of these, the associated preferencestructure using, to that end, tools analogous to those of Multidimensional Scaling.The resulting PS will be employed to extract knowledge for the subsequent negotiationprocesses and, should it be necessary, to determine the relative importance of thealternatives being compared using anyone of the existing procedures

Norm inequalities in multidimensional Lorentz spaces

In this paper we obtain necessary and sufficient conditions for double trigonometric series to belong to generalized Lorentz spaces, not symmetric in general. Estimates for the norms are given in terms of coefficients.

Lp theory for the multidimensional aggregation equation

"Vegeu el resum a l'inici del document del fitxer adjunt."

Power laws and scaling of rain events and dry spells in the Catalonia region

We analyze the statistics of rain-event sizes, rain-event durations, and dry-spell durations in a network of 20 rain gauges scattered in an area situated close to the NW Mediterranean coast. Power-law distributions emerge clearly for the dryspell durations, with an exponent around 1.50 ± 0.05, although for event sizes and durations the power-law ranges are rather limited, in some cases. Deviations from power-law behavior are attributed to finite-size effects. A scaling analysis helps to elucidate the situation, providing support for the existence of scale invariance in these distributions. It is remarkable that rain data of not very high resolution yield findings in agreement with self-organized critical phenomena.