Multiple Classification Method Research Articles

Mining the Arabidopsis thaliana genome for highly-divergent seven transmembrane receptors

A combination of multiple protein classification methods is described and used to identify a minimum set of 54 candidate seven transmembrane receptors in Arabidopsis thaliana.

Learning multicriteria fuzzy classification method PROAFTN from data

In this paper, we present a new methodology for learning parameters of multiple criteria classification method PROAFTN from data. There are numerous representations and techniques available for data mining, for example decision trees, rule bases, artificial neural networks, density estimation, regression and clustering. The PROAFTN method constitutes another approach for data mining. It belongs to the class of supervised learning algorithms and assigns membership degree of the alternatives to the classes. The PROAFTN method requires the elicitation of its parameters for the purpose of classification. Therefore, we need an automatic method that helps us to establish these parameters from the given data with minimum classification errors. Here, we propose variable neighborhood search metaheuristic for getting these parameters. The performances of the newly proposed method were evaluated using 10 cross validation technique. The results are compared with those obtained by other classification methods previously reported on the same data. It appears that the solutions of substantially better quality are obtained with proposed method than with these former ones.

A new multi-expert decision combination algorithm and its application to the detection of circumscribed masses in digital mammograms

A new multiple expert fusion algorithm is introduced, designated the “augmented behaviour-knowledge space method”. Most existing multiple expert classification methods rely on a large training dataset in order to be properly utilised. The proposed method effectively overcomes this problem as it exploits the confidence levels of the decisions of each classifier. It will be shown that this new approach is advantageous when small datasets are available, and this is illustrated in its application to the detection of circumscribed masses in digital mammograms, with very encouraging results.

High-SNR asymptotics for signal-subspace methods in sinusoidal frequency estimation

High-SNR-limit second-order properties of multiple signal classification (MUSIC), minimum-norm (MN), and subspace rotation (SUR) signal-subspace methods for sinusoidal frequency estimation are discussed. An alternative to large-sample analysis of the methods is presented. The two most important variants of these methods are considered in connection with the choice of the sample covariance matrix: the simpler technique follows the principle of a linear prediction, and the more complex one is based on the idea of a forward-backward prediction. Explicit expressions for the high-SNR covariance elements of the estimation errors associated with all the methods are derived. The expressions for the covariances are used to analyze and compare the statistical performances of MUSIC, MN, and SUR estimation methods in both of the variants, to discuss the problem of the optimal dimension of the data covariance matrix, and to study the limit statistical efficiency of the methods. Performances of the large-sample and high-SNR asymptotics derived using Monte Carlo simulations are presented. >

Music analysis applied to NMR

Here, we would like to present an alternative for this transform, the multiple signal identification and classification method (MUSIC). This is a spectral estimator, like FFT, that has proven to be very useful in molecular dynamics studies (3, 4), and other applications. For certain cases this method has advantages over the FFT

Multivariate analysis of the relationship of seven variables to blood pressure: Findings of the Chicago Heart Association Detection Project in Industry, 1967–1972

1. 1. As in the preceding cross-sectional study [1], four multivariate statistical methods —partial correlation, multiple cross classification, multiple logistical regression, and multiple linear regression—were used to evaluate the relationship between seven variables and blood pressure in over 21,000 employed men and women, white and black, ages 25–44 and 45–64 surveyed in almost 100 Chicago companies and organizations. The analyses were carried out separately for eight age-sex-race groups. The seven variables were relative weight, resting heart rate, plasma glucose one hr after 50 g oral load, serum uric acid and cholesterol, current cigarette smoking habit, and age. 2. 2. Consistent with the findings of the preceding report, the first three of these variables—relative weight, heart rate, and plasma glucose—were independently related to blood pressure with a high degree of consistency, with p values for statistical significance ≤0.01 or ≤0.001 in the great majority of analyses. 3. 3. Serum uric acid—a variable not evaluated in the preceding report—was also independently related to blood pressure in a great majority of the analyses with p values of ≤0.01 or ≤0.001. 4. 4. The findings with respect to serum cholesterol and blood pressure were generally negative, except for white males, an intriguing finding in view of a similar result from the preceding study. 5. 5. No positive relationship was found between cigarette smoking and blood pressure. 6. 6. Even within the narrow age bands studied—ages 25–44 and 45–64—age was significantly related to blood pressure in a great majority of analyses ( p values ≤0.01 or ≤0.001), independent of the six other variables. 7. 7. As in the preceding paper, when multivariate regression equations and their coefficients, computed from this experience of an entire age-sex-race group, were used to calculate an expectation of elevated blood pressure for each person, and then persons were ordered from low to high in expectation, a high proportion of all persons with recorded elevations of blood pressure were in the highest decile and quintile of expected prevalence—e.g. for persons age 25–44 in the four sex-race groups, from 45 to 56 per cent of observed cases with diastolic pressure ≥95 mm Hg. were in the highest quintile (20 per cent) of expected prevalence. The observed prevalence of elevated blood pressure was many times greater for the highest quintile of expected prevalence than for the lowest. A similar but less effective concentration of cases of elevated blood pressure was obtained with the multiple cross classification method, by dichotomizing five of the independent variables and identifying the substratum with any three, four or all five high.

Multivariate analysis of the relationship of six variables to blood pressure: Findings from Chicago Community Surveys, 1965–1971

1. 1. Four multivariate statistical methods—partial correlation, multiple cross classification, multiple logistic regression, and multiple linear regression—were used to study the relationship between six variables and blood pressure in over 13,000 men and women, white and black, ages 30–44 and 45–64 surveyed in two low-income Chicago communities. The analyses were carried out separately for eight age-sex-race groups. The six variables were relative weight, resting heart rate, plasma glucose 1 hr after 100 g oral load, hematocrit, serum cholesterol and age. 2. 2. The first three of these variables—relative weight, heart rate, and plasma glucose —were independently related to blood pressure with a high degree of consistency, with p values for statistical significance ≤0.001. 3. 3. Hematocrit was independently and additively related to diastolic—but not to systolic—pressure for the four age-race groups of women, but only for one of the four age-race groups of men. 4. 4. The findings with respect to serum cholesterol and blood pressure were generally negative. 5. 5. Even within the narrow age groups studied—ages 30–44 and 45–64—age was significantly related to systolic level particularly ( p values ≤0.01 or ≤0.001), independent of the five other variables. 6. 6. When multivariate regression equations and their coefficients, computed from the experience of an entire age-sex-race group, were used to calculate an expectation of elevated blood pressure for each person, and then persons were ordered from low to high in expectation, a high proportion of all persons with recorded elevations of blood pressure were in the highest decile and quintile of expected prevalence—e.g. for persons age 30–44 in the four sex-race groups, from 47–69 per cent of observed cases with diastolic pressure ≥95 mm Hg were in the highest quintile (20 per cent) of expected prevalence. The observed prevalence of elevated blood pressure was as much as 20 times as high for the highest quintile of expected prevalence as for the lowest. A similar but less effective concentration of cases of elevated blood pressure was obtained with the multiple cross classification method, by dichotomizing four of the independent variables and identifying the substratum with any two, three or all four high.