Volodymyr Lytvynenko

Inductive model of data clustering based on the agglomerative hierarchical algorithm

Model of data clustering system based on the complex use of agglomerative hierarchical algorithm and inductive modeling methods of complex systems is presented in the paper. The quality of clustering is evaluated by two equal power subsets with the use of complex balance criterion, which takes into account both the displacement the mass centers of the appropriate clusters of different subsets and distribution of objects in the appropriate clusters relative to the mass center. Evaluating the effectiveness of the proposed model was performed using data sets containing clusters of different shapes. Charts of the external and internal complex criterion values against clustering level were created, which allows to determine the optimal clustering of a data set.

The problem of system fault-tolerance

Abstract. System level self-diagnosis (SLSD) has been deeply investigated in literature. It aims at diagnosing systems composed by units, which are required to be able to test each other by exchanging information through available links. The article describes a simplified state-transition diagram model which gives a general impression of how checking, diagnosis and recovery can “conjointly” influence the system reliability and fault-tolerance. The model uses the integrated parameters and is very useful as a starting point and is a basis for further refinements.

Criterial analysis of gene expression sequences to create the objective clustering inductive technology

The paper presents the researches to determine the effectiveness of different criteria to estimate the complex biology objects clustering quality. The gene expression sequences of cancer patients were used as experimental data. The degree of the studied objects similarity was estimated by the comparison of the gene expression sequences profile using different metrics to estimate the objects proximity. The studies have shown that the best separating ability is obtained by using the correlation metric proximity of objects. Herewith the use of the CH criterion (Calinski-Harabasz) allows to get the most objective objects clustering by using simulated data. The presented research is focused mainly on the inductive model of the objective clustering, where the objects clustering is carried out concurrently on the two equal power subsets. In this case, the final decision about the objects grouping is accepted using the two subsets basing both on the internal clustering quality criteria estimating and the minimum value of the external criterion of clustering similarity.

Objective Clustering Inductive Technology of Gene Expression Sequences Features

Technology of high dimensional data features objective clustering based on the methods of complex systems inductive modeling is presented in the paper. Architecture of the objective clustering inductive technology as a block diagram of step-by-step implementation of the objects clustering procedure was developed. Method of criterial evaluation of complex data clustering results using two equal power data subsets is proposed. Degree of clustering objectivity evaluates on the basis of complex use of internal and external criteria. Researches on the simulation results of the proposed technology based on the SOTA self-organizing clustering algorithm using the gene expression data obtained by DNA microarray analysis of patients with lung cancer GEOD-68571 Array Express database, the datasets “Compound” and “Aggregation” of the Computing School of the Eastern Finland University and the data “seeds” are presented.

Model of the Objective Clustering Inductive Technology of Gene Expression Profiles Based on SOTA and DBSCAN Clustering Algorithms

The paper presents the hybrid model of the objective clustering inductive technology based on complex using of the self-organizing SOTA and the density DBSCAN clustering algorithms. The inductive methods of complex systems analysis were used as the basis to implement the objective clustering inductive technology of gene expression profiles. To estimate the clustering quality for equal power subsets (include the same quantity of pairwise similar objects) the complex multiplicative criterion was calculated as the combination of the Calinski-Harabasz criterion and WB-index. The external clustering quality criterion is calculated as the normalized difference of the internal clustering quality criteria for the equal power subsets. The final decision concerning the determination of the optimal parameters of the clustering algorithm operation is done based on the maximum value of the Harrington desirability function that takes into account both the character of the objects and the clusters distribution in various clustering and the difference between clustering, which are implemented on the equal power subsets. The studied data grouping within the framework of the objective clustering inductive technology was performed in two stages. Firstly, the studied gene expression profiles were grouped with the use DBSCAN clustering algorithm. Then, the obtained set of gene expression profiles was divided into two clusters using SOTA clustering algorithm. This step-by-step procedure of the data clustering crates the conditions to save more useful information for following data processing.

Implementation of the objective clustering inductive technology based on DBSCAN clustering algorithm

The paper presents the results of the research of the clustering algorithm DBSCAN practical implementation within the framework of the objective clustering inductive technology. As experimental, the data Aggregation and Compound of the Computing school of the East Finland University and the gene expression sequences of lung cancer patients of the database ArrayExpres were used. The architecture of the objective clustering model has been developed. The implementation of the model involves the parallel data clustering on the two equal power subsets, which include the same quantity of pairwise similar objects. The choice of the solution about parameters of the algorithm determination has been carried out based on the minimum value of the external clustering quality criterion, which calculated as normalized difference of the internal clustering quality criteria for the two subsets.

Multifractal spectra classification of flame luminosity waveforms

The article presents the results of multifractal and statistical analysis changes in the characteristics of the multifractal spectra of a burner flame luminosity waveforms in industrial boilers at different loads and air supply levels to obtain a characteristic space for solving classification problems.

A Fuzzy Model for Gene Expression Profiles Reducing Based on the Complex Use of Statistical Criteria and Shannon Entropy

The paper presents the technology of gene expression profiles reducing based on the complex use of fuzzy logic methods, statistical criteria and Shannon entropy. Simulation of the reducing process has been performed with the use of gene expression profiles of lung cancer patients. The variance and the average absolute value were changed within the defined range from the minimum to the maximum value, and Shannon entropy from the maximum to the minimum value during the simulation process. 311 gene expression profiles from 7129 were removed as non-informativity during simulation process. The structural block diagram of the step-by-step data processing in order to remove non-informativity gene expression profiles has been proposed as the simulation results.

Synthesis and learning of fuzzy neural networks for solving forecasting problems

Two variations of the network are described i.e. Larsen model and Tsukamoto model. In each case an artificial immune network is proposed as a means of building fuzzy network structure and performing parameters inference.

Evaluation of testing assignment for system level self-diagnosis

The paper concerns system level self-diagnosis (SLSD). SLSD aims at diagnosing systems composed by units with the requirement that they are able to test each other by exchanging information through available links. At this level of diagnosis, each particular test is considered as atomic. It means that the details of a test are abstracted (not considered), and only the result of test is taken into consideration. One of the main issues of SLSD is the issue of testing assignment that defines the possible set of tests among the system units. System testing assignment relies and depends on physical connections among the system units. The issue of testing assignment is tightly connected with the diagnosability problem of SLSD. Diagnosability problem of SLSD is the problem of how to determine the family of fault sets that a given testing assignment can diagnose for some fault model. In the paper, we have shown how different testing assignments can be evaluated and compared. For this, we suggest to use characteristic numbers. We also have shown how these characteristic numbers can be computed.