Neural networks and statistical analysis for classification of corneal videokeratography maps based on Zernike coefficients: a quantitative comparison

PURPOSE: The main goal of this study was to develop and compare two different techniques for classification of specific types of corneal shapes when Zernike coefficients are used as inputs. A feed-forward artificial Neural Network (NN) and discriminant analysis (DA) techniques were used. METHODS: The inputs both for the NN and DA were the first 15 standard Zernike coefficients for 80 previously classified corneal elevation data files from an Eyesys System 2000 Videokeratograph (VK), installed at the Departamento de Oftalmologia of the Escola Paulista de Medicina, São Paulo. The NN had 5 output neurons which were associated with 5 typical corneal shapes: keratoconus, with-the-rule astigmatism, against-the-rule astigmatism, "regular" or "normal" shape and post-PRK. RESULTS: The NN and DA responses were statistically analyzed in terms of precision ([true positive+true negative]/total number of cases). Mean overall results for all cases for the NN and DA techniques were, respectively, 94% and 84.8%. CONCLUSION: Although we used a relatively small database, results obtained in the present study indicate that Zernike polynomials as descriptors of corneal shape may be a reliable parameter as input data for diagnostic automation of VK maps, using either NN or DA.

Further psychometric study of the Beck Anxiety Inventory including factorial analysis and social anxiety disorder screening

Objective. To analyze the psychometric properties of the Brazilian Portuguese version of the Beck Anxiety Inventory (BAI) in terms of its internal consistency, scores distribution, concurrent and discriminant validity, and factorial analysis in a sample of university students and social anxiety disorder (SAD) cases and non-cases. Methods. A sample of Brazilian university students from the general population (N = 2314) and a sample of university students identified as cases (N = 88) and non-cases (N = 90) of SAD were assessed, using as a parameter the Structured Clinical Interview for the DSM-IV. The different instruments were completed individually in the presence of an experienced rater. Results. The BAI showed adequate internal consistency (0.88-0.92) and discriminant validity, with 0.74 sensitivity and 0.71 specificity for a cut-off score of 10. The factorial analysis suggested a three-factor solution to be the most adequate. Conclusions. The version of the BAI studied is quite adequate to be used in the context of Brazilian university students, identifying the presence of anxiety indicators. However, its usefulness to screen for SAD seems limited.

Comparative analysis between statistical and artificial intelligence models in business failure prediction

A growing number of predicting corporate failure models has emerged since 60s. Economic and social consequences of business failure can be dramatic, thus it is not surprise that the issue has been of growing interest in academic research as well as in business context. The main purpose of this study is to compare the predictive ability of five developed models based on three statistical techniques (Discriminant Analysis, Logit and Probit) and two models based on Artificial Intelligence (Neural Networks and Rough Sets). The five models were employed to a dataset of 420 non-bankrupt firms and 125 bankrupt firms belonging to the textile and clothing industry, over the period 2003–09. Results show that all the models performed well, with an overall correct classification level higher than 90%, and a type II error always less than 2%. The type I error increases as we move away from the year prior to failure. Our models contribute to the discussion of corporate financial distress causes. Moreover it can be used to assist decisions of creditors, investors and auditors. Additionally, this research can be of great contribution to devisers of national economic policies that aim to reduce industrial unemployment.

Comparative Analysis between Statistical and Artificial Intelligence Models in Business Failure Prediction

A growing number of predicting corporate failure models has emerged since 60s. Economic and social consequences of business failure can be dramatic, thus it is not surprise that the issue has been of growing interest in academic research as well as in business context. The main purpose of this study is to compare the predictive ability of five developed models based on three statistical techniques (Discriminant Analysis, Logit and Probit) and two models based on Artificial Intelligence (Neural Networks and Rough Sets). The five models were employed to a dataset of 420 non-bankrupt firms and 125 bankrupt firms belonging to the textile and clothing industry, over the period 2003–09. Results show that all the models performed well, with an overall correct classification level higher than 90%, and a type II error always less than 2%. The type I error increases as we move away from the year prior to failure. Our models contribute to the discussion of corporate financial distress causes. Moreover it can be used to assist decisions of creditors, investors and auditors. Additionally, this research can be of great contribution to devisers of national economic policies that aim to reduce industrial unemployment.

Multi-elemental analysis of ready-to-eat “baby leaf” vegetables using microwave digestion and high-resolution continuum source atomic absorption spectrometry

The mineral content (phosphorous (P), potassium (K), sodium (Na), calcium (Ca), magnesium (Mg), iron (Fe), manganese (Mn), zinc (Zn) and copper (Cu)) of eight ready-to-eat baby leaf vegetables was determined. The samples were subjected to microwave-assisted digestion and the minerals were quantified by High-Resolution Continuum Source Atomic Absorption Spectrometry (HR-CS-AAS) with flame and electrothermal atomisation. The methods were optimised and validated producing low LOQs, good repeatability and linearity, and recoveries, ranging from 91% to 110% for the minerals analysed. Phosphorous was determined by a standard colorimetric method. The accuracy of the method was checked by analysing a certified reference material; results were in agreement with the quantified value. The samples had a high content of potassium and calcium, but the principal mineral was iron. The mineral content was stable during storage and baby leaf vegetables could represent a good source of minerals in a balanced diet. A linear discriminant analysis was performed to compare the mineral profile obtained and showed, as expected, that the mineral content was similar between samples from the same family. The Linear Discriminant Analysis was able to discriminate different samples based on their mineral profile.

Optimization of River Water Quality Surveys by Multivariate Analysis of Physicochemical, Bacteriological and Ecotoxicological Data

This study aims to optimize the water quality monitoring of a polluted watercourse (Leça River, Portugal) through the principal component analysis (PCA) and cluster analysis (CA). These statistical methodologies were applied to physicochemical, bacteriological and ecotoxicological data (with the marine bacterium Vibrio fischeri and the green alga Chlorella vulgaris) obtained with the analysis of water samples monthly collected at seven monitoring sites and during five campaigns (February, May, June, August, and September 2006). The results of some variables were assigned to water quality classes according to national guidelines. Chemical and bacteriological quality data led to classify Leça River water quality as “bad” or “very bad”. PCA and CA identified monitoring sites with similar pollution pattern, giving to site 1 (located in the upstream stretch of the river) a distinct feature from all other sampling sites downstream. Ecotoxicity results corroborated this classification thus revealing differences in space and time. The present study includes not only physical, chemical and bacteriological but also ecotoxicological parameters, which broadens new perspectives in river water characterization. Moreover, the application of PCA and CA is very useful to optimize water quality monitoring networks, defining the minimum number of sites and their location. Thus, these tools can support appropriate management decisions.

Differentiation and genetic analysis of Rhodnius prolixus and Rhodnius colombiensis by rDNA and RAPD amplification

Domiciliated Rhodnius prolixus and sylvatic R. colombiensis were analyzed in order to confirm their genetic divergence and verify the risk that the latter represents in the domiciliation process, and to provide tools for identifying the sources of possible reinfestation by triatomines in human dwellings allowing control programs to be undertaken. Comparison of random amplified polymorphic DNA amplification patterns and cluster analysis suggests reproductive discontinuity between the two species. The calculated statistical F value of 0.24 and effective migration rate of 0.6 individuals per generation are insufficient to maintain genetic homogeneity between them and confirm the absence of present genetic flow. R. colombiensis presents higher intrapopulation variability. Polymerase chain reaction of ribosomal DNA supports these findings. The low genetic flow between the two species implies that R. colombiensis do not represent an epidemiological risk for the domiciliary transmission of Trypanosoma cruzi in the Tolima Department. The lower variability of the domiciliated R. prolixus could result in greater susceptibility to the use of pesticides in control programs.

Analysis of antenal sensilla patterns of Rhodnius prolixus from Colombia and Venezuela

Antennal sensilla patterns were used to analyze population variation of domestic Rhodnius prolixus from six departments and states representing three biogeographical regions of Colombia and Venezuela. Discriminant analysis of the patterns of mechanoreceptors and of three types of chemoreceptors on the pedicel and flagellar segments showed clear differentiation between R. prolixus populations east and west of the Andean Cordillera. The distribution of thick and thin-walled trichoids on the second flagellar segment also showed correlation with latitude, but this was not seen in the patterns of other sensilla. The results of the sensilla patterns appear to be reflecting biogeographic features or population isolation rather than characters associated with different habitats and lend support to the idea that domestic R. prolixus originated in the eastern region of the Andes.

Multivariate analysis and geostatistics of the fertility of a humic rhodic hapludox under coffee cultivation

The spatial variability of soil and plant properties exerts great influence on the yeld of agricultural crops. This study analyzed the spatial variability of the fertility of a Humic Rhodic Hapludox with Arabic coffee, using principal component analysis, cluster analysis and geostatistics in combination. The experiment was carried out in an area under Coffea arabica L., variety Catucai 20/15 - 479. The soil was sampled at a depth 0.20 m, at 50 points of a sampling grid. The following chemical properties were determined: P, K+, Ca2+, Mg2+, Na+, S, Al3+, pH, H + Al, SB, t, T, V, m, OM, Na saturation index (SSI), remaining phosphorus (P-rem), and micronutrients (Zn, Fe, Mn, Cu and B). The data were analyzed with descriptive statistics, followed by principal component and cluster analyses. Geostatistics were used to check and quantify the degree of spatial dependence of properties, represented by principal components. The principal component analysis allowed a dimensional reduction of the problem, providing interpretable components, with little information loss. Despite the characteristic information loss of principal component analysis, the combination of this technique with geostatistical analysis was efficient for the quantification and determination of the structure of spatial dependence of soil fertility. In general, the availability of soil mineral nutrients was low and the levels of acidity and exchangeable Al were high.

On the Schoenberg transformations in data analysis: theory and illustrations

The class of Schoenberg transformations, embedding Euclidean distances into higher dimensional Euclidean spaces, is presented, and derived from theorems on positive definite and conditionally negative definite matrices. Original results on the arc lengths, angles and curvature of the transformations are proposed, and visualized on artificial data sets by classical multidimensional scaling. A distance-based discriminant algorithm and a robust multidimensional centroid estimate illustrate the theory, closely connected to the Gaussian kernels of Machine Learning.

Characterization and classification of pequi trees (Caryocar brasiliense Camb.) based on the profile of volatile constituents using headspace solid-phase microextraction - gas chromatography - mass spectrometry and multivariate analysis

In order to determine the variability of pequi tree (Caryocar brasiliense Camb.) populations, volatile compounds from fruits of eighteen trees representing five populations were extracted by headspace solid-phase microextraction and analyzed by gas chromatography-mass spectrometry. Seventy-seven compounds were identified, including esters, hydrocarbons, terpenoids, ketones, lactones, and alcohols. Several compounds had not been previously reported in the pequi fruit. The amount of total volatile compounds and the individual compound contents varied between plants. The volatile profile enabled the differentiation of all of the eighteen plants, indicating that there is a characteristic profile in terms of their origin. The use of Principal Component Analysis and Cluster Analysis enabled the establishment of markers (dendrolasin, ethyl octanoate, ethyl 2-octenoate and β-cis-ocimene) that discriminated among the pequi trees. According to the Cluster Analysis, the plants were classified into three main clusters, and four other plants showed a tendency to isolation. The results from multivariate analysis did not always group plants from the same population together, indicating that there is greater variability within the populations than between pequi tree populations.

Analysis of intrinsic cardiac neuron activity in relation to neurogenic atrial fibrillation and vagal stimulation

La fibrillation auriculaire est le trouble du rythme le plus fréquent chez l'homme. Elle conduit souvent à de graves complications telles que l'insuffisance cardiaque et les accidents vasculaires cérébraux. Un mécanisme neurogène de la fibrillation auriculaire mis en évidence. L'induction de tachyarythmie par stimulation du nerf médiastinal a été proposée comme modèle pour étudier la fibrillation auriculaire neurogène. Dans cette thèse, nous avons étudié l'activité des neurones cardiaques intrinsèques et leurs interactions à l'intérieur des plexus ganglionnaires de l'oreillette droite dans un modèle canin de la fibrillation auriculaire neurogène. Ces activités ont été enregistrées par un réseau multicanal de microélectrodes empalé dans le plexus ganglionnaire de l'oreillette droite. L'enregistrement de l'activité neuronale a été effectué continument sur une période de près de 4 heures comprenant différentes interventions vasculaires (occlusion de l'aorte, de la veine cave inférieure, puis de l'artère coronaire descendante antérieure gauche), des stimuli mécaniques (toucher de l'oreillette ou du ventricule) et électriques (stimulation du nerf vague ou des ganglions stellaires) ainsi que des épisodes induits de fibrillation auriculaire. L'identification et la classification neuronale ont été effectuées en utilisant l'analyse en composantes principales et le partitionnement de données (cluster analysis) dans le logiciel Spike2. Une nouvelle méthode basée sur l'analyse en composante principale est proposée pour annuler l'activité auriculaire superposée sur le signal neuronal et ainsi augmenter la précision de l'identification de la réponse neuronale et de la classification. En se basant sur la réponse neuronale, nous avons défini des sous-types de neurones (afférent, efférent et les neurones des circuits locaux). Leur activité liée à différents facteurs de stress nous ont permis de fournir une description plus détaillée du système nerveux cardiaque intrinsèque. La majorité des neurones enregistrés ont réagi à des épisodes de fibrillation auriculaire en devenant plus actifs. Cette hyperactivité des neurones cardiaques intrinsèques suggère que le contrôle de cette activité pourrait aider à prévenir la fibrillation auriculaire neurogène. Puisque la stimulation à basse intensité du nerf vague affaiblit l'activité neuronale cardiaque intrinsèque (en particulier pour les neurones afférents et convergents des circuits locaux), nous avons examiné si cette intervention pouvait être appliquée comme thérapie pour la fibrillation auriculaire. Nos résultats montrent que la stimulation du nerf vague droit a été en mesure d'atténuer la fibrillation auriculaire dans 12 des 16 cas malgré un effet pro-arythmique défavorable dans 1 des 16 cas. L'action protective a diminué au fil du temps et est devenue inefficace après ~ 40 minutes après 3 minutes de stimulation du nerf vague.

Morphometric analysis and taxonomic revision of the Vriesea paraibica complex (Bromeliaceae)

The species related to Vriesea paraibica (Bromeliaceae, Tillandsioideae) have controversial taxonomic limits. For several decades, this group has been identified in herbarium collections as V. x morreniana, an artificial hybrid that does not grow in natural habitats. The aim of this study was to assess the morphological variation in the V. paraibica complex through morphometric analyses of natural populations. Two sets of analyses were performed: the first involved six natural populations (G1) and the second was carried out on taxa that emerged from the first analysis, but using material from herbarium collections (G2). Univariate ANOVA was used, as well as discriminant analysis of 16 morphometric variables in G1 and 18 in G2. The results of the analyses of the two groups were similar and led to the selection of diagnostic traits of four species. Lengths of the lower and median floral bracts were significant for the separation of red and yellow floral bracts. Vriesea paraibica and V. interrogatoria have red bracts; these two species are differentiated by the widths of the lower and median portions of the inflorescence and by scape length. These structures are larger in the former and smaller in the latter. Of the species with yellow floral bracts, V. eltoniana is distinguished by longer leaf blades and scapes and V. flava is characterized by its shorter sepal lengths. (C) 2009 The Linnean Society of London, Botanical Journal of the Linnean Society, 2009, 159, 163-181.

Main obstacles in the business environment affecting firm growth: a comparative analysis between Chinese and Brazilian SMEs

Usando dados do Banco Mundial, esta dissertação analisa as semelhanças e diferenças entre os obstáculos enfrentados pelas micro, pequenas e médias empresas (MSME) no Brasil e na China. Realizamos tanto uma análise comparativa entre os dos países como uma análise de clusters para confrontar os dados subjetivos e objetivos recolhidos, de modo a identificar o desempenho das empresas, levando em consideração as características dos países e o ambiente de negócios.

Multivariate analysis for the classification and differentiation of Madeira wines according to main grape varieties

In order to differentiate and characterize Madeira wines according to main grape varieties, the volatile composition (higher alcohols, fatty acids, ethyl esters and carbonyl compounds) was determined for 36 monovarietal Madeira wine samples elaborated from Boal, Malvazia, Sercial and Verdelho white grape varieties. The study was carried out by headspace solid-phase microextraction technique (HS-SPME), in dynamic mode, coupled with gas chromatography–mass spectrometry (GC–MS). Corrected peak area data for 42 analytes from the above mentioned chemical groups was used for statistical purposes. Principal component analysis (PCA) was applied in order to determine the main sources of variability present in the data sets and to establish the relation between samples (objects) and volatile compounds (variables). The data obtained by GC–MS shows that the most important contributions to the differentiation of Boal wines are benzyl alcohol and (E)-hex-3-en-1-ol. Ethyl octadecanoate, (Z)-hex-3-en-1-ol and benzoic acid are the major contributions in Malvazia wines and 2-methylpropan-1-ol is associated to Sercial wines. Verdelho wines are most correlated with 5-(ethoxymethyl)-furfural, nonanone and cis-9-ethyldecenoate. A 96.4% of prediction ability was obtained by the application of stepwise linear discriminant analysis (SLDA) using the 19 variables that maximise the variance of the initial data set.