Neural Network Cluster Analysis Research Articles

Нейросетевой кластерный анализ устойчивого развития экономики регионов Российской Федерации в аспекте принципов ESG

The sustainable development of the innovative economy in the regions of the Russian Federation is researched. From this position, the changes in the commissioning of capacities for the protection of water resources and atmospheric air from pollution, as well as the renewal of forests in Russia for the period 2017–2022 are presented. Taking into account the importance of responsible investment, investments in fixed assets aimed at environmental protection and rational use of natural resources of the Russian Federation in 2022 are given. The relevance of the research is due to the progress in the achievements of science and technology, as well as the existing strong competition. This requires the promotion of creative concepts in the management of sustainable socio-economic development of Russia. Such concepts may be based on consideration of ESG criteria: Environment — environment, Social — social development, Governance — corporate governance. In order to reduce the heterogeneity of the socio-ecological development of Russia, the multifactorial task of strategic management on the formation of a methodology based on ESG principles is analyzed. The authors propose a promising and productive method for solving multifactorial problems — neural network cluster analysis using information technology. It allowed us to distribute the regions of the Russian Federation into six cluster formations characterized by different levels of development in the perspective of ESG features according to a set of 11 indicators. The structure of clusters detailing the entry of Russian regions into them is given. The average values of the studied indicators of sustainable innovative development of the regional economy in each of the clusters and in the whole country are shown. It follows from the results of the work that the management’s focus on the synthesis of ESG principles and the productivity of responsible investment will become important components in the formation of strategies for the innovative development of the Russian economy. This will be one of the key factors in strengthening the economic security of the Russian Federation.

Quantitative identification and prediction of mixed lithology, Bohai Sea, China

The Paleogene Shahejie Formation in the KL16 oilfield, Bohai bay, is characterized by a thinly interbedded mixed sedimentary system, with complex sedimentary facies, lithologic types and distributions. It is hard for conventional logging methods to identify the lithology therein. In order to solve the difficulty in lithologic identification of mixed sedimentary system, analyses based on graph data base using elemental capture energy spectrum log have been proposed. Due to the different composition for the various minerals, we innovatively established the molar numbers of silicon, calcium, magnesium, and aluminum as characteristic parameters for sandstone, limestone, dolomite, and mudstone, and a graph clustering analysis method was applied to identify lithology. Considering the seismic waveforms corresponding to lithologic impedance of reservoir, three seismic phases were identified by neural network clustering analysis of seismic waveform, and the seismic attributes with high sensitivity to reservoir thickness were then selected to realize the fine description of the mixed carbonate-siliciclastic reservoir. Drilling results confirmed that the sedimentary facies were accurately identified, with reservoir prediction accuracy reaching up to 80%. Under the guidance of reservoir research, the oil-in-place discovered in the oilfield were estimated to be more than 5 million tonnes. This technology provides reference for the exploration and development of oilfields of mixed sedimentary system.

Improving the temporal resolution of event-based electron detectors using neural network cluster analysis

Novel event-based electron detector platforms provide an avenue to extend the temporal resolution of electron microscopy into the ultrafast domain. Here, we characterize the timing accuracy of a detector based on a TimePix3 architecture using femtosecond electron pulse trains as a reference. With a large dataset of event clusters triggered by individual incident electrons, a neural network is trained to predict the electron arrival time. Corrected timings of event clusters show a temporal resolution of 2 ns, a 1.6-fold improvement over cluster-averaged timings. This method is applicable to other fast electron detectors down to sub-nanosecond temporal resolutions, offering a promising solution to enhance the precision of electron timing for various electron microscopy applications.

The Use of Neural Network Modeling Methods to Determine Regional Threshold Values of Hydrochemical Indicators in the Environmental Monitoring System of Waterbodies.

The regulation of the anthropogenic load on waterbodies is carried out based on water quality standards that are determined using the threshold values of hydrochemical indicators. These applied standards should be defined both geographically and differentially, taking into account the regional specifics of the formation of surface water compositions. However, there is currently no unified approach to defining these regional standards. It is, therefore. appropriate to develop regional water quality standards utilizing modern technologies for the mathematical purpose of methods analysis using both experimental data sources and information system technologies. As suggested by the use of sets of chemical analysis and neural network cluster analysis, both methods of analysis and an expert assessment could identify surface water types as well as define the official regional threshold values of hydrochemical system indicators, to improve the adequacy of assessments and ensure the mathematical justification of developed standards. The process for testing the proposed approach was carried out, using the surface water resource objects in the territory of the Republic of Tatarstan as our example, in addition to using the results of long-term systematic measurements of informative hydrochemical indicators. In the first stage, typing was performed on surface waters using the neural network clustering method. Clustering was performed based on sets of determined hydrochemical parameters in Kohonen's self-organizing neural network. To assess the uniformity of data, groups in each of the selected clusters were represented by specialists in this subject area's region. To determine the regional threshold values of hydrochemical indicators, statistical data for the corresponding clusters were calculated, and the ranges of these values were used. The results of testing this proposed approach allowed us to recommend it for identifying surface water types, as well as to define the threshold values of hydrochemical indicators in the territory of any region with different surface water compositions.

Comparison of hierarchical clustering and neural network clustering: an analysis on precision dominance

A comparison of neural network clustering (NNC) and hierarchical clustering (HC) is conducted to assess computing dominance of two machine learning (ML) methods for classifying a populous data of large number of variables into clusters. An accurate clustering disposition is imperative to investigate assembly-influence of predictors on a system over a course of time. Moreover, categorically designated representation of variables can assist in scaling down a wide data without loss of essential system knowledge. For NNC, a self-organizing map (SOM)-training was used on a local aqua system to learn distribution and topology of variables in an input space. Ternary features of SOM; sample hits, neighbouring weight distances and weight planes were investigated to institute an optical inference of system’s structural attributes. For HC, constitutional partitioning of the data was executed through a coupled dissimilarity-linkage matrix operation. The validation of this approach was established through a higher value of cophenetic coefficient. Additionally, an HC-feature of stem-division was used to determine cluster boundaries. SOM visuals reported two locations’ samples for remarkable concentration analogy and presence of 4 extremely out of range concentration parameter from among 16 samples. NNC analysis also demonstrated that singular conduct of 18 independent components over a period of time can be comparably inquired through aggregate influence of 6 clusters containing these components. However, a precise number of 7 clusters was retrieved through HC analysis for segmentation of the system. Composing elements of each cluster were also distinctly provided. It is concluded that simultaneous categorization of system’s predictors (water components) and inputs (locations) through NNC and HC is valid to the precision probability of 0.8, as compared to data segmentation conducted with either of the methods exclusively. It is also established that cluster genesis through combined HC’s linkage and dissimilarity algorithms and NNC is more reliable than individual optical assessment of NNC, where varying a map size in SOM will alter the association of inputs’ weights to neurons, providing a new consolidation of clusters.

Identification of pathobiomechanical markers of statokinesiograms on the example of neural network identification of a post-stroke state

The purpose of this study is neural network modeling and determination of the parameters of statokinesiograms, which are carriers of useful information about the features of postural regulation, which determined the obtained trajectory of movements of the human center of mass. A technique for obtaining informative markers by identifying clustering centroids based on self-organizing Kohonen neural networks with the Euclidean metric has been developed. Kohonen networks trained without a teacher (that is, without the use of a priori diagnostic information about the state of the subjects) are a powerful and informative method that allows you to obtain not only graphic, but also mathematical markers of health disorders in subjects, to explore the biomechanics of micromovements of the center of pressure based on a flexible mathematical apparatus – neural network cluster analysis. On the example of the identification of the post-stroke state, a neural network analysis of stabilometric data was carried out and a method for identifying pathobiomechanical markers of statokinesiograms was shown, which made it possible to standardize the type affiliation of the groups of subjects. With the help of neural networks, it was possible to identify clusters that are amenable to biomedical interpretation with a reliability of up to 95.9 %, which determines the theoretical significance of the results obtained. Decision trees and a supervised multilayer neural network were also considered. A multilayer neural net-work made it possible to identify markers of health disorders with a probability of 71.9 %.

Application of Kohonen Neural Network in Sports Cluster

People’s physical fitness is directly linked to the national physique of a country. It is an important analysis indicator of the country’s comprehensive national strength and economic level. Moreover, students’ physical and mental health is in a stage of rapid development. Their physical health is not only directly related to their study and life at this stage but it will also have a profound impact on the physical level, health status, and work ability of adults. Starting from the cluster analysis of sports, this article explores the communication effects of sports, communication strategies, and the relationship between sports and the development of users’ healthy habits by defining concepts and types and combining quantitative and qualitative analysis. Study the relationship and then analyze whether sports applications bring people the effect of promoting healthy behavior. This article retrieved documents about the application of Kohonen neural network in sports cluster analysis in domestic literature databases such as Weipu, Wanfang data, and CNKI. A total of 144 documents were retrieved from the database, and the retrieved documents were collected for sports activities. The study of class analysis can avoid human subjective factors and obtain clustering results quickly and objectively, thus providing an ideal clustering method for comprehensive evaluation of sports. The experiment proves the cluster analysis of the impact of sports on people’s physical fitness, the heterogeneity test results are 0% (boys) and 3% (girls), it is believed that there is no statistical heterogeneity in the physical fitness of middle school boys in each study, and P < 0.001 , indicating that the influence of sports on people’s physical fitness is significantly different between the experimental group and the control group. This shows that the application of Kohonen neural network clustering analysis method has great practical value for comprehensively evaluating people’s physical functions and physical fitness. It is an objective, reasonable, effective, and rapid quantitative evaluation method.

Ring structures of the Norilsk metallogenic district in neural network cluster analysis of geophysical data

An integrated geophysical and mathematical model of the Norilsk metallogenic region has been built according to the authors’ method of automatic express interpretation of areal geophysical data. It consists of fifteen levels of hierarchy. The model is based on the use of the mathematical apparatus of fuzzy logic — artificial neural networks without a teacher using the method of self-organizing Kohonen maps. Formation of cluster groups is justified. Clusters characterize to the greatest extent possible connections between multidimensional geophysical data. The presence of relationships between them is analyzed by identifying correlation dependences. Analysis of various geophysical transformants using self-organizing Kohonen maps is carried out. A number of input indicators-representatives of their groups are determined, on the basis of which a geophysical-mathematical model is built. A model in the form of a two-dimensional map of clusters using fuzzy logic tools. Terminological sets were formed for each group of clusters. The form of membership functions of previously unknown geological objects is given according to new interpreted data. The calculated clusters according to the results of the study reflect the northwestern fragment of the basaltic magmatism fields of the trap formation, within which two ring structures have been distinguished. In the western part of the study area, the model ring structure reflects the Bolgokhtokhsky stock of granodiorites. The second, the Pyasinskaya ring structure, previously unknown, indicates the presence of a dome-shaped object at a great depth — a possible source of root zones of intrusions of the Norilsk complex. The conducted experimental study confirmed the adequacy of the constructed model and the effectiveness of its use for the purpose of express analysis of geophysical data and decision-making in geological prospecting tasks.

Analysis of Topics of Interdisciplinary Research in the Field of Library Science and Information Science from the Perspective of Altmetrics

On the basis of identifying the interdisciplinary research topics in the field of Library and Information Science, this paper constructed Altmetrics indicator systems and makes SOFM neural network clustering analysis on the comprehensive influence of the research topics, in order to provide reference for the interdisciplinary crossover and integration of Library and Information discipline in China under the background of “New Liberal Arts”. The study found that the 20 interdisciplinary research topics could be divided into four types according to their influence composition characteristics: comprehensive hotspots, online social hotspots, academic hotspots, potential hotspots. Interdisciplinary hot topics were important for the intersection and integration of library and information field and other disciplines, and high impact research topics deserved close attention from Chinese scholars.

Research of the innovative development of the Russian Federation regions and its impact on the eco-friendliness of the economy based on neural network cluster analysis for the purpose of economic security

This article is devoted to the assessment of innovative development of the Russian Federation regions using self-organizing Kohonen maps. The relevance of this paper is due to the need to develop various criteria for making decisions on state support for regions, the use of which will allow to conduct a comprehensive assessment of regional development. Its findings should serve as a reliable, high-quality source of information about the feasibility and type of support for a particular region. The study provides a justification for a set of indicators that are used in conducting cluster analysis. The methodology developed by the authors provides an opportunity to divide the regions into clusters, conduct their quantitative and qualitative analysis, and further predict the key indicators of the regions based on their belonging to different clusters. All this allows to study regional innovative development in dynamics. The authors draw a conclusion about the uneven innovative development of the Russian regions and the ambiguity of connections between the indicators of the share of organizations engaged in technological innovations and the degree of environmental friendliness of the economy. The findings of this study will be useful to public authorities when making decisions on the allocation of budgetary funds for the development of territories.

A Note on Comparison between Statistical Cluster and Neural Network Cluster

Background: Cluster analysis is a data reduction technique in rows of the data matrix. This technique is widely used in engineering, biology, society, pattern recognition, and image processing. Objective: In this paper, self organized map (SOM) using the artificial neural network and different statistical techniques of cluster analysis are used on Population data of 33 districts of Rajasthan with 9 variables for comparison purpose. Methods: The goal of this work is to identify the most suitable technique for clustering the data by using the artificial neural network and different statistical clustering techniques. We received all patents regarding artificial neural network and k-means cluster method. Conclusion: The k-means cluster analysis is found as good as Neural Network cluster analysis, whereas Hierarchical cluster analysis and two steps cluster analysis provide some variation from the neural network cluster analysis.

Relationships between putative G-quadruplex-forming sequences, RecQ helicases, and transcription.

BackgroundPutative G-quadruplex-forming sequences (PQS) have long been implicated in regulation of transcription, though the actual mechanisms are not well understood. One proposed mechanism involves the activity of PQS-specific helicases belonging to the RecQ helicase family. However, patterns of PQS that correlate with transcriptional sensitivity to RecQ helicases are not well studied, and no adequate transcriptional model exists to account for PQS effects.MethodsTo better understand PQS transcriptional effects, we analyze PQS motifs in genes differentially-transcribed in Bloom Syndrome (BS) and Werner Syndrome (WS), two disorders resulting in loss of PQS-interacting RecQ helicases. We also correlate PQS genome-wide with transcription in multiple human cells lines while controlling for epigenetic status. Finally, we perform neural network clustering of PQS motifs to assess whether certain motifs are over-represented in genes sensitive to RecQ helicase loss.ResultsBy analyzing PQS motifs in promoters of genes differentially-transcribed in BS and WS, we demonstrate that abundance of promoter PQS is generally higher in down-regulated genes and lower in up-regulated genes, and show that these effects are position-dependent. To interpret these correlations we determined genome-wide PQS correlations with transcription while controlling for epigenetic status. Our results identify multiple discrete transcription start site-proximal positions where PQS are correlated with either increased or decreased transcription. Finally, we report neural network clustering analysis of PQS motifs demonstrating that genes differentially-expressed in BS and WS are significantly biased in PQS motif composition.ConclusionsOur findings unveil unappreciated detail in the relationship between PQS, RecQ helicases, and transcription. We show that promoter PQS are generally correlated with reduced gene expression, and that this effect is relieved by RecQ helicases. We also show that PQS at certain positions on the downstream sense strand are correlated with increased transcription. We therefore propose a new transcriptional model in which promoter PQS have at least two distinct types of transcriptional regulatory effects.Electronic supplementary materialThe online version of this article (doi:10.1186/s12881-015-0236-4) contains supplementary material, which is available to authorized users.

Near‐surface properties of an active fault derived by joint interpretation of different geophysical methods ‐ the Arava/Araba Fault in the Middle East

ABSTRACTThe motion of tectonic plates is accommodated at fault zones. One of the unanswered questions about fault zones relates to the role they play in controlling shallow and local hydrology. This study focuses on the Arava/Araba Fault (AF) zone, the southern portion of the Dead Sea Transform (DST) in the Middle East. We combine seismic and electromagnetic methods (EM) to image the geometry and map the petro‐physical properties and water occurrence in the top 100 m of this active fault. For three profiles, P‐velocity and resistivity images were derived independently. Using a neural network cluster analysis three classes with similar P‐velocity and resistivities could then be determined from these images. These classes correspond to spatial domains of specific material and wetness. The first class occurs primarily east of the fault consisting of ‐wet‐ sand (dunes) and brecciated sediments, whereas the second class composed of similar material located west of the fault is ‐dry‐. The third class lies at depth below ca. 50 m and is composed of highly deformed and weathered Precambrian rocks that constitute the multi‐branch fault zone of the AF at this location. The combination of two independent measurements like seismics and EM linked by a stringent mathematical approach has thus shown the potential to delineate the interplay of lithology and water near active faults.