Use Of Discriminant Functions Research Articles

НЕОДНОРОДНОСТЬ ТЕЧЕНИЯ ОСТРОГО ПЕРИОДА ГЕМОРРАГИЧЕСКОГО ИНСУЛЬТА: АНАЛИЗ ПРИ ПОМОЩИ МЕТОДОВ МНОГОМЕРНОЙ СТАТИСТИКИ

Background. Hemorrhagic stroke is an important medical and social problem both in the world and in the Russian Federation due to high parameters of morbidity, mortality and disability. Aim. To compare expert assessments and formalized multivariate statistical procedures in analysis of clinical inhomogeneity of patients with intracranial hematoma of supratentorial location. Materials and Methods. 75 Patients who took treatment in the neurovascular department of Ryazan Regional Clinical Hospital with the diagnosis of hemorrhagic stroke, were examined. Of them, there were 40 men and 35 women with the mean age 68.1 years. Results. Primarily, on the basis of expert assessments, the patients were divided to 2 groups: with unfavorable course and with relatively favorable course – satisfactory condition, regress of symptoms, recovery of the level of consciousness. In the first stage, in result of the primary expert assessment, the patients were divided to subgroups with favorable and unfavorable course. After that, significant for selection of groups variables were used in selection of clusters: method of hierarchial tree for determination of the number of groups; k-means method for identification of their elements. Discrimination analysis was used for selection of variables in NIHSS and GCS, and also for assessment of canonical correlations. After that, cluster analysis of NIHSS and GCS dynamics was conducted on the 1st, 3rd, 21st day for selection of groups in high-dimensional space of signs with exclusion of subjective expert assessments. Three main groups of patients were selected. In the second stage, in accordance with the number of groups, patients belonging to the corresponding clusters were identified using the k-means method. Cluster 1 included patients with a poor prognosis, clusters 2 and 3 suggested a more favorable course of the acute period with worse parameters in cluster 2. Conclusions. Use of discriminant functions confirms the role of severity of depression of consciousness and of the volume of hematoma in the unfavorable course.

Stream technologies of data analysis in organizational and technological design in construction

The issues of intellectual processing of large data arrays in construction are considered. The concept of an automated system for analyzing big data is proposed. The technology of data mining assumes the possibility of estimating the time and cost of performing work on the basis of the used organizational and technological models and real indicators of time and labor observed by these models, and, accordingly, the formation of specific technical and economic indicators. Such indicators are of great importance in the current conditions of activity of construction organizations, they are used by the management of construction organizations in the preparation of tender documents, the development of construction organization projects, and in settlements with subcontractors. The proposed sequence of actions when using data mining technology to estimate the time and cost of work and resource consumption will allow at the stage of a feasibility study or decision to conclude contractual obligations to provide a better understanding of future costs and timing of work. This technique assumes the possibility of its use by both data analysis specialists, data scientist and process engineers, BIM managers, and the like and has a number of advantages due to the use of discriminant functions for data analysis.

Stream Technologies Of Data Analysis In Construction Organizational And Technological Design

The issues of intellectual processing of large information data arrays in construction are considered. The concept of an automated system for analyzing big data is proposed. The technology of data mining assumes the possibility of estimating the time and cost of performing work on the basis of the used organizational and technological models and real indicators of time and labor observed by these models, and, accordingly, the formation of specific technical and economic indicators. Such indicators are of great importance in the current conditions of activity of construction organizations, they are used by the management of construction organizations in the preparation of tender documents, the development of construction organization projects, and in settlements with subcontractors. The proposed sequence of actions when using data mining technology to estimate the time and cost of work performance and resource consumption will make it possible at the stage of a feasibility study or decision to conclude contractual obligations to provide a better understanding of future costs and timing of work. This technique assumes the possibility of its use by both data analysis specialists, and process engineers, BIM managers, and other specialists and has a number of advantages due to the use of discriminant functions for data analysis.

Determination of adverse pregnancy outcomes using discriminant function analysis

The article is devoted to determination of adverse pregnancy outcomes, using discriminant function analysis, which improves the efficiency of diagnosis and prognosis of this condition by improving the accuracy of determining the likely outcome, which is achieved by obtaining informative, prognostic significant signs. The calculation of diagnostic coefficient values by the selected method allows revealing 6 significant characteristics of protein, carbohydrate, nucleic acid and mineral metabolism and canonical coefficients to determine the pregnancy outcome. The use of discriminant functions and the most significant characteristics allows synthesizing a new mathematical model for determining the pregnancy outcome, which allows describing the condition of the patient. The validity of the synthesized mathematical model has been confirmed by independent testing factors on sample examination with classification accuracy for which amounted to 95,8 %. This method has several advantages: takes into account the variability of parameters included in the model, considering the totality of all the prognostic value and indicates the proportion of the impact of each characteristic on the formation of the severity of the condition. The results provide a timely opportunity to choose the optimal treatment and care of women with threatened miscarriage.

Quality indices in degraded pasture in hilly relief

Devido a influencia do relevo na variabilidade do solo, condicoes pedogeneticas homogeneas tem sido recomendadas para desenvolvimento de indices de qualidade do solo baseados em seus atributos quimicos, fisicos e biologicos. Tal fato limita o desenvolvimento e aplicacao de um indice para um mesmo pedoambiente de relevo mais acidentado. O presente trabalho teve como objetivo avaliar a cobertura do solo e atributos quimicos e fisicos do solo como indicadores de qualidade relacionando-os com aspectos visuais de degradacao das pastagens e contextualizando a influencia do relevo. Para tanto, separaram-se niveis de degradacao a partir de aspectos visuais, avaliou-se a cobertura do solo, atributos quimicos e fisicos do solo. Tais niveis foram definidos de forma menos subjetiva a partir das funcoes discriminantes geradas utilizando-se informacoes da cobertura do solo. Pela Analise de Fatores observa-se que altitude e faces de exposicao ao sol estao relacionadas a explicacao de uma menor proporcao da variacao total dos dados. Em contrapartida, a declividade controla a maior parte da variacao dos atributos do solo, principalmente os fisicos que nao apresentaram uma relacao direta e clara com a cobertura do solo e com niveis de degradacao dificultando o uso desses atributos como indicadores de qualidade. Essa relacao com os atributos dos solos se diferencia principalmente em funcao da face de exposicao: na face Leste/Sul observa maior relacao dos atributos fisicos com a cobertura vegetal, diferentemente na face Oeste/Norte. De fato, a influencia do relevo no controle da intensidade de insolacao e da umidade do solo modifica a dinâmica da relacao solo-vegetacao. Assim, a mensuracao da cobertura vegetal e uso de funcoes discriminantes torna-se, em relevo mais acidentado, uma forma mais eficiente na discriminacao entre niveis de degradacao, ponderando adicionalmente a influencia de fatores do ambiente (temperatura, intensidade de insolacao, umidade do solo) na qualidade das pastagens.

The Ben Lawers Historic Landscape Project: Simultaneous Multi-element Analysis of Former Settlement and Arable Soils by X-ray Fluorescence Spectrometry

X-ray fluorescence spectrometry (XRFS) was used in the analysis of A horizon soil samples collected from a former farming settlement and its associated area of infield (i.e. arable) located in the Central Highlands of Scotland. To date, XRFS has not been extensively used in geoarchaeological research, but in our study the simultaneous multi-element capabilities of this instrumental technique allowed the total concentrations of 25 major, minor, and trace elements to be fully quantified with acceptable levels of accuracy and precision. Included within this group of chemical elements are a number (e.g. Ba, Ca, P, Pb, Sr and Zn) that have proved to be of value to archaeological interpretation in earlier investigations undertaken in Scotland. In our preliminary work documented here, significant differences were found between the A horizon soils of former settlement and infield areas for 18 chemical elements. Subjecting the XRFS data—and three other measured variables: soil organic carbon (SOC), pH and A horizon depth—to discriminant analysis indicates that soils of former settlement and arable farming can be effectively classified according to their pH, SOC content and Ca, Cu, Mg, Rb, and Zn concentrations. The inference is that areas of former infield and settlement elsewhere at this study location in the Central Highlands may be able to be identified according to their soil chemical composition and use of discriminant function, even though the surface remains of pre-eighteenth century settlement sites are not readily evident today because they were constructed of perishable materials.

Sex determination from the occipital condyle: Discriminant function analysis in an Eighteenth and Nineteenth Century British sample

Fragmentary human remains compromised by different types of inhumation, or physical insults such as explosions, fires, and mutilations may frustrate the use of traditional morphognostic sex determination methods. The basicranium is protected by a large soft tissue mass comprising muscle, tendon, and ligaments. As such, the occipital region may prove useful for sex identification in cases of significantly fragmented remains. The aims of this paper are to (1) evaluate sexual dimorphism in British cranial bases by manually recorded unilateral and bilateral condylar length and width as well as intercondylar measurements and (2) develop discriminant functions for sex determination for this cranial sample. The crania selected for this study are part of the 18th-19th century documented skeletal collection of St. Bride's Church, Fleet Street, London. Adult human skulls (n = 146; male75/female71) were measured to derive statistical functions. Results indicated that expression of sexual dimorphism in the occipital condylar region within the St. Bride's population is demonstrable but low. Crossvalidated classification accuracy ranged between 69.2 and 76.7%, and sex bias ranged from 0.3 to 9.7%. Therefore, the use of discriminant functions derived from occipital condyles, especially in British skeletal populations, should only be considered in cases of fragmented cranial bases when no other morphognostic or morphometric method can be utilized for sex determination.

Short Note: A method for sexing the chicks of Antarctic shags

Several studies have been carried out during the last fifteen years on reproduction, population dynamics and diet of the Antarctic shag, Phalacrocorax bransfieldensis Murphy, at different localities in the South Shetland Islands (reviewed in Casaux & Barrera-Oro 2006). In both the colonies studied and in other colonies counted we observed that the number of breeding pairs was steadily decreasing (Casaux & Barrera-Oro 2006). Casaux & Baroni (2002) had earlier suggested that such a decreasing trend might be related, at least partially, to a marked decrease in the inshore populations of two fish prey species, the marbled notothen Notothenia rossii Richardson and the humphead notothen Gobionotothen gibberifrons Lönnberg (Barrera-Oro et al. 2000), which had been studied over a period of 19 years in coastal waters of the South Shetland Islands. Exactly how a reduction in prey availability affects the shag populations (e.g. migration of breeders to other colonies in the area or to new breeding areas, a decrease in the rate of recruitment, an increase in adult mortality, variation in the age at first breeding, etc) is not clear. To investigate this, we started a banding programme at Nelson Island, South Shetland Islands. We postulated that the processes might operate with different intensities on individuals of different sexes, so all individuals in each population studied needed to be sexed. This posed problems for chicks which have monomorphic plumage and no differences in vocalisations (Casaux & Baroni 2000), so that the normal methods for sexing in the field would not work. As most of the external morphological characters in the chicks of Antarctic shags have stabilized by 45–50 days old (Casaux 1998), Casaux & Baroni (2000) had suggested that the use of discriminant functions originally developed for adults could be an appropriate method to sex chicks more than 50 days old.

Univariate sex dimorphism in the Nepalese dentition and the use of discriminant functions in gender assessment

Sex dimorphism in the Nepalese dentition is described using univariate and discriminant analyses. Canines showed the greatest univariate sex dimorphism, followed by the buccolingual (BL) dimension of maxillary first and second molars. Overall, the maxillary teeth and BL dimensions showed greater univariate sex differences. However, less than half of the measured variables (46.4%) showed statistically significant differences between the sexes and the magnitude of sex dimorphism was reduced when compared to other populations. Moreover, reverse dimorphism – where females showed larger teeth than males – was observed in the mesiodistal dimension of mandibular second premolars. This reflects reduction in sexual dimorphism observed through human evolution and the consequent overlap of tooth dimensions in modern males and females. A specific purpose of the study was to develop discriminant functions to facilitate sex classification. A group of functions were developed considering the possibility of missing teeth and/or jaws in forensic scenarios. The functions permitted moderate to high classification accuracy in sexing (67.9% using maxillary posterior teeth; 92.5% using teeth from both jaws). The superior expression of sex dimorphism by means of discriminant functions is in contrast to the univariate results. This is due to discriminant analysis utilising the inter-relationship between all teeth within a dentition—these tooth correlations are not utilised in univariate analysis which results in a loss of information. It is inferred that large-scale statistically significant univariate differences are not a prerequisite for sex assessment.

Preliminary chemometric study of minerals and trace elements in Spanish infant formulae

The concentrations of minerals (Na, K, P, Ca and Mg) and trace elements (Fe, Zn, Cu, Mn, Se, Al, Cd and Pb) in a total of 105 different infant formulae (starter, follow-up, premature, specialised and soya formulae) marketed in Spain were determined by atomic spectrometry (flame and electrothermal) and inductively coupled plasma emission spectroscopy after acid-microwave decomposition. On the basis of the elements distribution, a preliminary chemometric study with the use of pattern recognition methods was carried out. Hierarchical cluster analysis (HCA), principal component analysis (PCA), as unsupervised exploratory techniques, and linear discriminant analysis (LDA), were applied to characterise, classify and distinguish the different types of infant formulae. The HCA results showed that mineral and trace element content data support adequate information to obtain the infant formula differentiation. PCA permitted the reduction of 13 variables to four principal components accounting for 61.9% of the total variability. This four-factor model interprets reasonably well the correlations of these studied elements. The obtained element associations may be attributed to the composition of matrix ingredients, the contamination during elaboration, the additives and mineral supplements added and the present tendency of standardization in the manufacture of infant formulae. The application of LDA gave a 77.1% of infant formulae correctly assigned with three clearly differentiated and two overlapped groups. The use of discriminant functions, as a complementary tool, to distinguish the different types depending on protein matrix of infant formula, is also discussed. This survey shows that HCA, PCA and LDA techniques appear useful tools for the characterisation and classification of infant formulae using their elemental profile.

How to distinguish most similar insect species – improving the stereomicroscopic and mathematical evaluation of external characters by example of ants

Abstract: The non‐destructive investigation of type specimens, the need to interpret available taxonomic descriptions, and the requirement to give statistical evidence for a presented hypothesis will guarantee that careful investigation and numeric recording of phenotypic characters will remain the basis of taxonomic work in insects, even if the progress of molecular genetics should continue to unexpected levels. A process of how to discriminate most similar insect species by means of external morphology is explained step by step. The process includes: (a) the reproducible and practicable definition of characters; (b) the reduction of errors during stereomicroscopic investigation and data recording; (c) the application of pair comparisons in discriminative calculations; (d) the description and removal of allometric variance; (e) the use of discriminant functions; and (f) the final removal of variance by consideration of nest means. The avoidance or reduction of 10 types of error occurring during stereomicroscopic investigation is explained. A linear regression of a weighted (indexed) character against a measure of body size is the recommended method to describe continuous allometries. In the case of discontinuous allometries and/or if discrete worker castes are present, different regression lines are calculated with this method for up to four separate size intervals and then connected in the form of a single‐valued polygon. Two types of discriminant function, a linear function without consideration of character correlations and a function using the function values of normal distributions and considering character correlations, both proved to be powerful in separating most similar species. The whole procedure from the beginning of data recording to the final calculation can increase, in the most successful case, the portion of individuals outside the interspecific overlap range from 25 to 100%. In the opposite extreme, the procedure can increase the overlap and thus reveal pseudo‐discriminative characters or pseudo‐species that are artefacts of allometric growth.

Evaluation of an aspartame loading test for the detection of heterozygotes for classical phenylketonuria.

Classical phenylketonuria (PKU) is an inborn error of metabolism of autosomal recessive inheritance characterized by the accumulation of phenylalanine (Phe) in tissues due to Phe-4-hydroxylase deficiency. Several methods have been developed for the detection of PKU heterozygotes based on the determination of plasma Phe and tyrosine (Tyr) levels, on the analysis of the Phe/Tyr and Phe2/Tyr ratios and on the use of discriminant functions. The objective of the present study was to test the value of loading with aspartame (a sweetener consisting of Phe, aspartate and methanol) for the identification of PKU carriers. The study was conducted on 22 obligate heterozygotes and 27 controls. Two blood samples were collected (under fasting conditions and 30 min after the loading) for fluorometric determination of Phe and Tyr. Phe, Phe/Tyr and Phe2/Tyr values were higher in heterozygotes, whereas Tyr was higher in controls in both situations investigated. Linear discriminant function was considered to be the best parameter for differentiation of the individuals in the two groups. Under the conditions employed in the present study, aspartame loading did not show any advantages in discriminating between PKU carriers and normal individuals when compared to the same analysis performed under fasting conditions.

Sexing great SkuasCatharacta skuaby discriminant analysis using external measurements

A discriminant function used to sex Great Skuas at Foula, Shetland correctly classified 87.5% of the sample from which it was derived, and several lines of evidence suggest that it performed similarly with a sample of live birds caught over a three‐year period. There was no evidence that any of the variables used in the analysis changed during the period of the study or were affected by conditions at the colony during birds’ development. However, this may not be the case with birds caught in future years. Culmen length was correlated with age, and so was not suitable for inclusion in the analysis. The discriminant function derived by one worker performed comparatively poorly with a sample of birds measured by a different worker, despite the fact that their measurements were in close agreement with each other. We advocate great caution in the use of discriminant functions, particularly when a function has been derived by a different worker.

Climatic analysis of a phenological zonation: A multivariate approach

Phenological observations gathered from a network established in Québec since 1971 have been used to achieve a phenoclimatic regionalization of the agricultural territory. In order to verify that these zones reflect climatic differences and, if they do, to understand the nature of these differences, we have performed multivariate analyses combining phenological and meteorological data. Phenological zonation emphasizes a thermal gradient related to plant location. Thermal contrasts among zones are greatest in spring whereas equalization occurs during the summer. These results stress the importance of spring climatic conditions on plant adaptation in a temperate climate. Phenological classification of meteorological records with the use of discriminant functions presents a valuable approach to assess the influence of climatic variability on plant acclimation potential. This study confirms the complementarity assumption between climate and phenology and illustrates the usefulness of phenological information to characterize the bioclimatic potential of a territory.

Use of Discriminant Function to Characterize Ruffed Grouse Drumming Sites in Georgia: A Critique

Hale et al. (J. Wildl. Manage. 46:115123, 1982) used univariate statistical comparisons and discriminant function analysis to identify the most important characteristics of ruffed grouse (Bonasa umbellus) drumming sites in Georgia. Their univariate comparisons were backed by sound biological discussion. However, their use of discriminant function analysis on such a limited data set is not enlightening and may lead to incorrect biological conclusions.

Use of Discriminant Function to Characterize Ruffed Grouse Drumming Sites in Georgia: A Reply

Use of Discriminant Function to Characterize Ruffed Grouse Drumming Sites in Georgia: A Reply

Some Observations of the Use of Discriminant Analysis in Ecology

The application of discriminant analysis in ecological investigations is discussed. The appropriate statistical assumptions for discriminant are illustrated, and both classification and group separation approaches are outlined. Three assumptions that are crucial in ecological studies are discussed at length, and the consequences of their violation are developed. These assumptions are: (1) equality of dispersions, (2) identifiability of prior probabilities, and (3) precise and accurate estimation of means and dispersions. The use of discriminant functions for purposes of interpreting ecological relationships is also discussed. It is suggested that the common practice of imputing ecological meaning to the signs and magnitudes of coefficients be replaced by an assessment of structure coefficients. Finally, the potential and limitations of representation of data in canonical space are considered, and some cautionary points are made concerning ecological interpretation of patterns in canonical space.

Comparison of Multivariate Techniques in the Prediction of Juvenile Postparole Outcome

Several strategies for the prediction of institutional outcome have been attempted by researchers in juvenile and parole prediction. The majority of researchers have utilized a variation of multiple regression, while a few have made use of discriminant functions and factor analysis. The present study contrasted the three approaches in terms of predictive efficiency for juvenile recidivism. A total of 579 juvenile cases were selected from the Washington state juvenile archives during the calendar year of 1971. Forty predictive variables from personal histories and behavioral records of the Office of Juvenile Rehabilitation were collated and quantified for the study. The total sample was divided randomly into two groups for purposes of double cross-validation. On each of the two sample halves, a set of predictors of juvenile outcome were generated by each of three methods, factor analysis, stepwise regression, and stepwise discriminant functions. Validation of the three methods was assessed by (1) D2 or discriminability of each predictive variable on the cross-validation sample, and (2) the accuracy of classifying each juvenile based on discriminant analyses. The results showed clearly that stepwise discriminant functions provided the more consistent selection of variables with seven of 10 variables the same for all samples. Also, discriminant functions was shown to manifest the most predictive efficiency with significantly fewer misclassifications and greater discriminability. By contrast, stepwise regression and factor analysis as strategies manifested more misclassifications and less discriminability. In addition, the two methods were found to be unreliable in the selection of predictive variables for the two validation samples. Results were discussed in terms of the shrinkage problems of regression analyses and the homogeneity of variance associated with factor analysis.

Discriminant Functions, Role Orientations and Judicial Behavior: Theoretical and Methodological Linkages

IN A RECENT ARTICLE Sidney Ulmer presented an interesting theoretical use of discriminant functions, or fact pattern analysis.' Ulmer's research focused upon a concept which he termed and its influence upon judicial behavior. Political activism is defined as the ability of political or non-legal factors to explain variation in voting behavior. 2 His substantive application of fact pattern analysis concerned the behavior of five

Hybrid optical-digital radiography-based system for lung disease detection

A hybrid optical-digital system for automated measurement and disease classification of chest radiographs has been constructed. Initial results from the use of this system for classification of opacity textural patterns associated with a specific diffuse lung disease (coal workers pneumoconiosis) are presented. This system consists of three subsystems which include a digital image scanner for lung field detection, a computercontrolled X - Y transport for film positioning, and a Fraunhofer diffraction pattern sampling unit for opacity detection and measurement. A minicomputer is used to integrate these subsystems as well as render a film quality and disease classification through the use of discriminant functions and probability estimates.