Rule Of Classification Research Articles

META-MODEL OF DESIGN OF INFORMATION TECHNOLOGY PROTOTYPE OF CLASSIFICATION OF OBJECTS BY IMAGE SHAPE USING GMDH NEURAL NETWORKS IN THE NOTATION OF A UNIFIED MODELING LANGUAGE

The article is devoted to the development of intelligent monitoring systems of technological processes based on the use of machine vision systems. The principles of object-oriented formalization of the process of designing information technology of the classification of objects by the geometric shape of the image obtained from the machine vision system are proposed in the article. The proposed information technology is based on the use of machine learning technologies and provides for the selection of the best structure of the classifier model from a set of candidate models. Construction of candidate models is based on the use of the group method of data handling, based on the principles of self-organization of models. The outline of the image of the object obtained from the intelligent monitoring system is the input information. The set of input data contains a set of morphometric parameters that describe the geometric shape of the figure formed by the contour of the image of the object, as well as the label of the class to which the object belongs. The formation of the set of input data is implemented in the block «Image Processing». The decisive rule of classification is built in the block of synthesis of models of information technology of classification. GMDH neural networks were used as an algorithm for model synthesis. The choice of the best structure of the model is performed by a set of criteria. The information technology for constructing the classifier model is implemented by supplementing the block of algorithms for synthesis of MSUA models with the block «Class of models», which implements the process of selecting the class of functions for building models, and the block «Verification of models», which implements the best model structure. Construction of the meta-model of the design process was performed using a unified modeling language. The functional meta-model is represented by a use case diagram.

Entropy Based k Nearest Neighbor Pattern Classification (EbkNN): En-route to Achieving a High Accuracy in Breast Cancer Diagnosis

Entropy based k-Nearest Neighbor pattern classification (EbkNN) is a variation of the conventional k-Nearest Neighbor rule of pattern classification, which exclusively optimizes the value of k-neighbors for each test data based on the calculations of entropy. The formula for entropy used in EbkNN is the one that has been defined popularly in information theory for a set of n different types of information (class) attached to a total of m objects (data points) with each object defined by f features. In EbkNN that value of k is chosen for discrimination of given test data for which the entropy is the least non-zero value. Other rules of conventional kNN are retained in EbkNN. It is concluded that EbkNN works best for binary classification. It is computationally prohibitive to use EbkNN for discriminating the data points of the test dataset into number of classes greater than two. The biggest advantage of EbkNN vis-à-vis the conventional kNN is that in one single run of EbkNN algorithm we get optimum classification of test data. But conventional kNN algorithm has to be run separately for each of the selected range of values of k, and then the optimum k to be chosen from amongst them. We also tested our EbkNN method on WDBC (Wisconsin Diagnostic Breast Cancer) dataset. There are 569 instances in this dataset and we made a random choice of first 290 instances as training dataset and the rest 279 instances as test dataset. We got an exceptionally remarkable result with EbkNN method- accuracy close to 100% and better than the ones got by most of the other researchers who worked on WDBC dataset.

Ontology for preserving the knowledge base of traditional dances (OTD)

PurposeSystematic organization of domain knowledge has many advantages in archiving, sharing and retrieval of information. Ontologies provide a cushion for such practices in the semantic Web environment. This study aims to develop an ontology that can preserve the knowledge base of traditional dance practices.Design/methodology/approachIt is hypothesized that an ontology-based approach for the chosen domain might boost collaborative research prospects in the domain. A systematic methodology was developed for modeling the ontology based on the analytico-synthetic rule of library classification. Protégé 5.2 was used as an editor for the ontology using the Web ontology language combined with description logic axioms. Ontology was later implemented in a local GraphDB repository to run queries over it.FindingsThe developed ontology on traditional dances (OTD) was tested using the dances of the Rabha tribes of North East India. Rabha tribes are from an indigenous mongoloid community and have a robust presence in Southeast Asian countries, such as Myanmar, Thailand, Bangladesh, Bhutan and Nepal. The result from HermiT reasoner found the presence of no logical inconsistency in the ontology, while the OOPS! pitfall checker tool reported no major internal inconsistency. The induced knowledge base of traditional dances of the Rabha’s in the developed OTD was further validated based on some competency questions.Research limitations/implicationsIn the growing trend of globalization, preservation of the cultural knowledge base of human societies is an important issue. Traditional dances reflect a strong base of the cultural heritage of human societies as they are closely related to the lifestyle, habitat, religious practices and festivals of a specific community.Originality/valueThe current study is exclusively designed, keeping in mind the variables of traditional dance domain based on a survey of the user- and domain-specific needs. The ontology finds probable uses in traditional knowledge information systems, lifestyle-based e-commerce sites and e-learning platforms.

Nonparametric discrimination of areal functional data

We consider a new nonparametric rule of classification, inspired from the classical moving window rule, that allows for the classification of spatially dependent functional data containing some completely missing curves. We investigate the consistency of this classifier under mild conditions. The practical use of the classifier will be illustrated through simulation studies.

КЛАССИФИКАЦИЯ СОСТОЯНИЙ ОБЪЕКТА, ФУНКЦИОНИРОВАНИЕ КОТОРОГО ОПИСЫВАЕТСЯ ОДНОМЕРНОЙ АВТОРЕГРЕССИЕЙ

Рассмотрена задача классификации состояний динамического объекта, который может находиться в двух классах состояний. Функционирование объекта в классах описывается различными авторегрессионными моделями. Построено правило классификации и исследованы его свойства.

Classification of the state of the dynamic system, which is functioned by which is described by vector autoresis

Within the framework of the article, the problem of statistical classification of states of a dynamic system is solved, which can be in two classes of states, in each of which its operation is described by its own system of autoregressive equations with a priori unknown parameters. It is assumed that the following conditions are fulfilled: a) two classes of states are described by the same sets of observed input and output variables; b) the output variables, both in the first and in the second class, are determined by different sets of regressors (input variables); c) the models of functioning in the first and second classes are different both in terms of coefficients and in the structure of autoregressive models; d) the covariance matrices of random variables in the functioning models and the observation models for the first and second classes are different. The rule of classification is constructed and its properties are investigated.The experience of successfully solving problems of detecting changes in the properties of dynamic systems based on regression equations in the work, where an approach to constructing mathematical models for monitoring the technical condition of power and power plants in long-term operation was proposed, shows the feasibility of applying this approach to solving problems of controlling the operation of rocket-space objects technology.The problem of classifying states of a dynamic system, which can be in two classes of states, is considered. The functioning of the system in classes is described by various systems of autoregressive equations. The rule of classification is constructed and its properties are investigated.

Using mathematical criteria of evaluation for diagnostics results of cadets’ training in affective sphere

Objective: substantiation of the expediency of using mathematical evaluation criteria for determining the means of diagnosing a set of knowledge, skills and other competences acquired during the training of military men in the emotional (affective) sphere. Material: the study was attended by 67 cadets of the Air Force University, who are not the pilots and are at the 3rd year of study. For assessing the ability to manage their behavior through management of emotions, a technique of studying the emotional intelligence of military personnel at different stages of training and physical tests are related to overcoming objective and subjective obstacles. For the processing of the results was used the method of correlation, discriminant analysis and decision tree methods for the classification problem. The significance level for all statistics was established at p <0.05. Results: A positive linear relationship between the levels of performance of special exercises and the level of emotional stability were established. The models of discriminant analysis and the rule of classification of the recognition of the level of emotional stability according special physical preparedness level are determined. It was established that exercises in the obstacles course and turns in the gymnastics wheel are the main variables that allow classifying the students according to the level of emotional stability. Conclusions: The using of mathematical criteria for evaluating the results of the implementation of special exercises can differentiate according to the objective characteristics of cadets in category and simplifies the methodological work on recognition in the dynamics of physical improvement of military personnel, a local assessment of emotional and volitional resistance to the negative impact of the factors of modern combat.

Classification of autochthonous dengue virus type 1 strains circulating in Japan in 2014

In this paper, we classify by representativeness the elements of a set of complete genomic sequences of Dengue Virus Type 1 (DENV-1), corresponding to the outbreak in Japan during 2014. The set is coming from four regions: Chiba, Hyogo, Shizuoka and Tokyo. We consider this set as composed of independent samples coming from Markovian processes of finite order and finite alphabet. Under the assumption of the existence of a law that prevails in at least 50% of the samples of the set, we identify the sequences governed by the predominant law (see [1, 2]). The rule of classification is based on a local metric between samples, which tends to zero when we compare sequences of identical law and tends to infinity when comparing sequences with different laws. We found that the order of representativeness, from highest to lowest and according to the origin of the sequences is: Tokyo, Chiba, Hyogo, and Shizuoka. When comparing the Japanese sequences with their contemporaries from Asia, we find that the less representative sequence (from Shizuoka) is positioned in groups considerably far away from that which includes the sequences from the other regions in Japan, this offers evidence to suppose that the outbreak in Japan could be produced by more than one type of DENV-1.

The theory of nonlinear systems as an instrument for solving engineering problems

The article outlines theoretical, methodological and practical issues of modern control and optimization theory, as well as the problems of nonlinear systems theory. Theoretical conclusions and results allowed to build mathematical models applicable to the management of objects of different nature with different principles of action, in particular, to the management of complex technical and technological objects that can be considered as nonlinear dynamic systems. The authors find it appropriate to consider nonlinear dynamic integral models as Volterra integro-power series from many functional arguments with multidimensional weight functions and a certain finite set of inputs to the system. The set of multidimensional kernels of integral Volterra operators completely characterizes the nonlinear and dynamic properties, and, consequently, the technical state of the initial system. The application of Volterra series based models allows to take into account the nonlinear and inertial properties of the initial nonlinear dynamic system more fully and accurately, it also makes the model diagnostic of a technical system more universal, raises the reliability of the forecast. The diagnostic procedure in this case is aimed at defining Volterra kernels based on the data of “input-output” experiment and building the diagnostic system of attribute in the space of which the decisive rule of optimal classification is created.

On study designs and hypotheses for clinical trials with predictive biomarkers

Recent interest in conducting clinical trials with predictive biomarkers has generated research in comparing relative efficiency of different trial designs. We find these comparisons of efficiency mostly misleading since they are based on different hypotheses. In this paper, we discuss several commonly used trial designs and consider the hypotheses that each design is capable to address. We first consider the ideal situation of no classification errors, then the more realistic situation where marker assay's sensitivity, specificity and the rule of classification are imperfect. We pay special attention to the differences between treatment utility versus absolute treatment effect, and marker by treatment interaction versus marker utility.

On the consistency of a new kernel rule for spatially dependent data

We consider a new kernel rule of classification for spatially dependent data. This nonparametric rule allows for the classification of missing observations. We investigate the consistency of this classification rule and we propose a method for bandwidth selection.

Statistical Inference Based on Unbiased Estimators for Intervals of Random Length*

An estimator for a distribution quintile is proposed and studied with the use of the unbiased approach within the framework of the exponential distribution family model. The application of the quintile of the distribution function’s unbiased estimator to the construction of a chi-square test for the hypothesis about distribution type is justified. Limit behavior of the Bayes risk estimator for the optimal decision rule of group classification is studied under the assumption that the elements of training samples have distributions belonging to the same one-parameter exponential family.

Molecular dynamic simulation on the state of water in poly(vinyl alcohol) hydrogel

In this paper, a molecular dynamic model for PVA hydrogel crosslinked by radiation is proposed. The rule of classification and statistical methods of water molecules in PVA hydrogel is formulated. Through molecular dynamic simulation, the content of non-freezing water, freezing water and free water in various cross-linking degree can be determined. The effect of cross-linking degree on the content of non-freezing water is also revealed by the simulation results. The results of molecular dynamic simulation show good agreements with the experimental data. Especially, it was shown that in a specific range of cross-linking degree, the content of non-freezing water in PVA hydrogel remains constant.

On nonparametric classification for weakly dependent functional processes

The purpose of this paper is to investigate the moving window rule of classification to classify functions under mixing conditions. We consider a random variable X taking values in a metric space (ℱ ,ρ ) with label Y ∈ {0,1}. We extend some results on consistency and strong consistency of the moving window rule from the i.i.d. case to the weakly dependent case under mild assumptions. The practical use of the moving window rule will be illustrated through a simulation study. The performance of the moving window rule is investigated.

심전도 신호의 리듬 특징을 이용한 부정맥 검출

In this paper, we look into previous research in relation to each processing step for ECG diagnosis and propose detection and classification method of arrhythmia using rhythm features of ECG signal. Rhythm features for distribution of rhythm and heartbeat such as identity, regularity, etc. are extracted in feature extraction, and rhythm type is classified using rule-base constructed in advance for features of rhythm section in rhythm classification. Experimental results for all of rhythm types in the MIT-BIH arrhythmia database show detection performance of 100% for arrhythmia with only normal rhythm rule and applicability of classification for rhythm types with arrhythmia rhythm rules.

Oscillations of the band gap of single-walled carbon nanotubes in the range of very small diameters

The electronic properties of carbon nanotubes with controllable chirality indices are studied in the range of very small diameters (0.3–2.0 nm) of semiconductor single-walled carbon nanotubes. The density functional theory (DFT) in the local density approximation (LDA) (DFT-LDA) method and Gaussian and TubeGen software are used in the study. An unsteady oscillatory variation in the band gap is established. Such behavior is due to the small chirality indices that define the symmetry and curvature of nanotubes. It is found that, for semiconductor nanotubes with very small diameters and (0, 4) and (0, 5) chirality indices, the band structure is degenerate and this degeneracy is responsible for the metal properties and violation of the rule of the 3k classification of single-walled nanotubes according to their electrical properties.

Feature Extraction Based on Maximum Nearest Subspace Margin Criterion

Based on the classification rule of sparse representation-based classification (SRC) and linear regression classification (LRC), we propose the maximum nearest subspace margin criterion for feature extraction. The proposed method can be seen as a preprocessing step of SRC and LRC. By maximizing the inter-class reconstruction error and minimizing the intra-class reconstruction error simultaneously, the proposed method significantly improves the performances of SRC and LRC. Compared with linear discriminant analysis, the proposed method avoids the small sample size problem and can extract more features. Moreover, we extend LRC to overcome the potential singular problem. The experimental results on the extended Yale B (YALE-B), AR, PolyU finger knuckle print and the CENPARMI handwritten numeral databases demonstrate the effectiveness of the proposed method.

C-Erb-b2 Oncogene Expression in Intraductal Proliferative Lesions of the Breast

The ductal intraepithelial neoplasia (DIN) classification which proposes new approaches to the diagnosis, terminology and differential diagnosis of intraductal proliferative lesions of the breast was applied to a series of female patients comprising C-erbB2 oncogene expression which may serve as an adjunct to the morphology by immunohistochemistry. The study was performed using the data of 94 patients. There was no difficulty encountered in the diagnosis of intraductal hyperplasia (IDH). In patients with Atypical Ductal Hyperplasia (AIDH), the diagnosis could be made by using the 2-mm rule of the DIN classification in patients who exhibited cytologic and structural characteristics of Ductal Carcinoma in Situ (DCIS) alone or in conjunction with classical IDH patterns. However, in lesions that mimicked classical IDH patterns despite displaying cytological features of in situ carcinomas, the experience and view point of the pathologist played a more prominent role. When the DIN classification criteria were applied to grade DCIS lesions, although the system was found to be practical, it did not provide adequate differentiation in intermediate grade (grade II-DIN 2) patients and further improvement was considered desirable. Fourty-five cases (47.8%) IDH, 19 (20.2%) AIDH, and 30 (31.9%) were DCIS. There were statistically significant differences in the levels of c-erbB2 oncogene expression between IDH, AIDH and DCIS lesions (p<0.001). In DCISs, grade, cell size, pleomorphic nuclear atypia showed statistically significant associations with c-erbB2 oncogene expression. These results suggest that c-erbB2 oncogene expression is a valuable marker in the differential diagnosis and prognostic evaluation of patients with intraductal proliferative lesions.

Sparse linear discriminant analysis by thresholding for high dimensional data

In many social, economical, biological and medical studies, one objective is to classify a subject into one of several classes based on a set of variables observed from the subject. Because the probability distribution of the variables is usually unknown, the rule of classification is constructed using a training sample. The well-known linear discriminant analysis (LDA) works well for the situation where the number of variables used for classification is much smaller than the training sample size. Because of the advance in technologies, modern statistical studies often face classification problems with the number of variables much larger than the sample size, and the LDA may perform poorly. We explore when and why the LDA has poor performance and propose a sparse LDA that is asymptotically optimal under some sparsity conditions on the unknown parameters. For illustration of application, we discuss an example of classifying human cancer into two classes of leukemia based on a set of 7,129 genes and a training sample of size 72. A simulation is also conducted to check the performance of the proposed method.

Hierarchical subdivision of the Cenozoic Era: A venerable solution, and a critique of current proposals

If Tertiary and Quaternary are to be formally defined and ranked, then only one topology and two ranking options for the major subdivisions of the Cenozoic Era are consistent with the principles of hierarchical classification. These are: 1) for the Tertiary and Quaternary to be ranked as Suberas (with the Paleogene and Neogene ranked as Periods of the Tertiary), or 2) for the Tertiary and Quaternary to be ranked as Periods (with the Paleogene and Neogene ranked as Subperiods of the Tertiary). The first option would leave the Period rank empty for Quaternary time, however, so the second option is preferred here. In neither case is an extension of the Neogene to the present permitted. If desired, however, “Early Cenozoic” and “Late Cenozoic” could be defined as formal unranked chrons, with the latter term providing an unambiguous expression for the same concept as the extended Neogene. Attempts to justify an extended Neogene by reference to a statement arising from the 1948 International Geological Congress fail because they are based on a misinterpretation of that statement. Likewise, an appeal to the alleged original meaning of the Neogene fails because it is factually superficial and would also have unacceptable consequences if applied as a general principle to the rest of the time scale. Four arguments against a ranked Tertiary and/or Quaternary involve the status of these names as parts of an obsolete classification; the disparity in their durations; an analogy between them and the unranked Precambrian; and their supposed ambiguity. These arguments are readily refuted. Arguments in favor of a ranked Tertiary and Quaternary include the continuing widespread use of these names, their practical relevance for geologic maps, and their role in honoring the pioneers of stratigraphy. A recent recommendation by the International Commission on Stratigraphy (ICS) would decouple the beginning of the Quaternary from the beginning of the Pleistocene, have the Neogene and Pliocene overlap the Tertiary/Quaternary boundary, and have the Gelasian Age belong to both the Pliocene and the Quaternary. These results violate the existing Guidelines of the ICS. This same proposal would also regard the Quaternary as a Subera within the Neogene Period (the latter being extended to the present). However, this arrangement violates the most fundamental rule of hierarchical classification in that a unit of superior rank would be included entirely within a unit of inferior rank. Given these problems, the International Union of Geological Sciences (IUGS) must reject the ICS recommendation. An improved version of the ICS proposal (with the Quaternary ranked as a Period and the Paleogene and extended Neogene ranked as Suberas) would still prohibit the Tertiary from being ranked and would either violate “Simpson's rule” or else require a new name for the traditional concept of the Neogene. Because the classification of the Cenozoic advocated here is strictly hierarchical, violates none of the existing Guidelines of the ICS, obeys Simpson's rule, requires no new names, allows the Tertiary, Quaternary, Paleogene, and Neogene to be formally ranked, and has been in general use for more than a century, it is to be preferred. Only two current options for defining the Quaternary and Pleistocene are valid, namely, to leave the beginnings of these units where they are now (Vrica GSSP at 1.8 Ma), or to move both to the Piacenzian/Gelasian boundary (Monte San Nicola GSSP at 2.6 Ma). The latter option is not prohibited by Lyell's “types” of the Older Pliocene nor (after 2008) by the existing Guidelines of the ICS, but if enacted could jeopardize the stability of the rest of the standard global time scale.