Re-evaluating compute performance in SBC clusters: HPL benchmarking across generations

The formulation in this research is whether school culture, teaching facilities, motivation and job satisfaction simultaneously affect the performance of teachers in the elementary School (SD) of cluster state I Se-District Stabat, Langkat District, and the purpose of this research is to know and analyze whether school culture, teaching facilities, motivation and work satisfaction simultaneously affect the performance of teachers in the elementary School (SD) state cluster I Se-District Stabat, Langkat District. This research is a quantitative descriptive study. Data collection techniques are conducted through interviews, questionaire questions and documentation studies. Sample in this study as many as 76 teachers at elementary School (SD) Negeri cluster I Se-District Stabat district of Langkat. Variables are measured at Likert scale. Hypothesis testing using multiple linear regression analyses through F-test and T-Test. The results of the test in unison showed that the variables of the school culture, work facilities and motivation and job satisfaction as a free variable (X) have significant effect on the teachers ' performance variables of the state elementary School cluster I Se-District Stabat District, Langkat. Partially, the school culture has no effect on the state teachers ' performance of cluster I Se-District Stabat, Langkat District. Furthermore, the work facility has a positive and significant effect on the state elementary school teachers cluster I Se-district Stabat in Langkat District. The motivation has positive and significant impact on the state elementary school teachers cluster I Se-district Stabat in Langkat District. Job satisfaction positively and significantly affect the teacher's performance of State Elementary school cluster I Se-District Stabat Langkat District

Evaluation of web service clustering using Dirichlet Multinomial Mixture model based approach for Dimensionality Reduction in service representation

This chapter explores the role of social capital in shaping inter-firms relations within local clusters and identifies whether a lack of social capital can be considered an impediment to cluster formation and development in post-communist countries. It is important to note that despite offering a definition, this chapter does not provide one model of social capital, nor define one type of impact on cluster performance. Social capital is one element among many other determinants and studying the link between social capital and cluster performance does not mean asserting that social capital is a positive value per se for clusters. However, attention is focused on some major features that characterise social capital and that positively impact on business clusters development: a sound base of trust among economic and institutional actors, together with valued and acknowledged co-operation. The chapter is structured as follows: First, parallel definitions of social capital and clusters are provided; second, the links between the two concepts are analysed (in particular the impact of social capital on cluster building and performance); third, specific issues to postcommunist countries are raised; and lastly, a policy debate is initiated.

This article aims to determine and analyse the main approaches of evaluation of cluster or group performance under information asymmetry within cluster. It is assumed that one of the most relevant causes of information asymmetry inside the business clusters are the different interests of its stakeholders and their willingness to dominate. This paper contributes to the further analysis, development and generalisation of evaluation approach of cluster performance and of the impact of information asymmetry on the activities of business clusters. Purpose – is to investigate the impact of information asymmetry on the evaluation on performance of business clusters and the methods and approaches of evaluation of performance with regard to information asymmetry. Design/methodology/approach – general overview of research papers presenting concepts and methodologies of evaluation of performance with regard to information asymmetry. Findings – information asymmetry has a significant impact on the performance of business clusters, and can be the decisive factor in the viability of a cluster. The members of cluster can be seem as subjects willing to dominate in cluster and to gain a relatively more portion of profit of clusters, some conflicts of interests can appear. However, there is no universal approach for evaluation of the impact of information asymmetry on cluster or group efficiency. This paper aims to highlight the main types of information asymmetry and respective approaches of evaluation analysed by the researchers.Research limitations/implications – the complexity and nature of information that can be used in the process of creation of innovation. The strong assumptions on information asymmetry from one side and the lack of advanced investigation focussed on the evaluation of business clusters performance efficiency under information asymmetry from other side are the most relevant limitations of research. Therefore the conclusions are focussed only on the conceptual level and analysis of possible further steps in creation of respective methods or models. Practical implications – information asymmetry has a significant impact on the activities and performance of business clusters, and can be the decisive factor for a viability of a cluster and creation of innovations. This study will contribute to the further development and generalisation of evaluation approach of cluster performance and of the impact of information asymmetry on the activities of business clusters. Originality/Value – This case in terms of business cluster performance and creation of innovations is not exhaustively analysed by other researchers. This paper is one of the first attempts to describe and make an assessment of the evaluation of clusters with financial contagion in the Baltic States. The findings of this article should ground the further steps of the creation of evaluation of performance efficiency under information asymmetry.

This paper introduces an affordable, scalable, and hands-on series of projects for teaching parallel computing and networking concepts to undergraduate STEM students. A cluster computer composed of Raspberry Pis is presented along with a proposed design concept and basic instructions for configuring the cluster. Use cases are presented for exploring the performance of the Pi cluster and examining the consequences of unbalanced task distribution across the cluster. The performance of the cluster is tested using both simple numerical integration and adaptive integration methods. In the case of simple integration, results show that the cluster provides speedup in accordance with expectations due to the equal time-complexity of individual computations. Adaptive integration serves as an educational point about the importance of equitable task management across the cluster since the tasks assigned to individual threads may be of different time-complexities. This project is suitable for advanced undergraduate STEM majors and can be tailored to the preferences and goals of the course or instructor.

The phenomenon we observe is that some teachers remain reluctant to make teaching preparations, such as creating syllabi and lesson plans, and there is a less harmonious relationship between teachers and their colleagues. Therefore, the principal is rumored to be less capable of having a good leadership style and creating a conducive work climate. The principal is less able to improve the work spirit and performance of teachers. This study aims to determine the influence of the principal’s leadership style, school work climate, and teacher work spirit, both partially and simultaneously, on teacher performance in Cluster IV, Sawan District. This type of research is quantitative descriptive research. This study uses a survey assessment approach. The data collection technique in survey research uses an instrument in the form of a questionnaire. The data analysis technique used was Multiple linear regression. The results of the study showed that 1) There is a determination between the principal’s leadership style and teacher performance in Cluster IV, Sawan District, 2) There is a determinant between the school work climate and teacher performance in Cluster IV, Sawan District, 3) There is a determinant between teacher work enthusiasm and teacher performance in Cluster IV, Sawan District, 4) There is a joint determinant between the principal’s leadership style, school work climate and work enthusiasm with teacher performance in Cluster IV, Sawan District. The coefficient of determination is 0.523 or 52.3%.

Bridging capital and performance in clustered firms: The heterogeneous effect of knowledge strategy

Despite an increasing consensus regarding the significance of properly identifying the most suitable clustering method for a given problem, a surprising amount of educational research, including both educational data mining (EDM) and learning analytics (LA), neglects this critical task. This shortcoming could in many cases have a negative impact on the prediction power of both the EDM and LA based approaches. To address such issues, this work proposes an evaluation approach that automatically compares several clustering methods using multiple internal and external performance measures on 9 real-world educational datasets of different sizes, created from the University of Tartu's Moodle system, to produce two-way clustering. Moreover, to investigate the possible effect of normalization on the performance of the clustering algorithms, this work performs the same experiment on a normalized version of the datasets. Since such an exhaustive evaluation includes multiple criteria, the proposed approach employs a multiple criteria decision-making method (i.e., TOPSIS) to rank the most suitable methods for each dataset. Our results reveal that the proposed approach can automatically compare the performance of the clustering methods and accordingly recommend the most suitable method for each dataset. Furthermore, our results show that in both normalized and nonnormalized datasets of different sizes with 10 features, DBSCAN and k-medoids are the best clustering methods, whereas agglomerative and spectral methods appear to be among the most stable and highly performing clustering methods for such datasets with 15 features. Regarding datasets with more than 15 features, OPTICS is among the top-ranked algorithms among the nonnormalized datasets, and k-medoids is the best among the normalized datasets. Interestingly, our findings reveal that normalization may have a negative effect on the performance of certain methods, e.g., spectral clustering and OPTICS; however, it appears to mostly have a positive impact on all of the other clustering methods.

Network based determinants of innovation performance in yacht building clusters

As distributed storage clusters have been used more and more widely in recent years, data replication management has become a hot research topic. In storage clusters, internal network bandwidth is usually a scarce resource. Misplaced replicas may take up too much network bandwidth and greatly deteriorate the overall performance of the cluster. Based on multi-objective evolutionary algorithm(MOEA), we developed a replication management scheme to improve performance by reducing internal network traffic and balancing load of storage clusters. The replica placement problem is formulated as some multi-objective programming optimization problems and solved by MOEA to get a Pareto solution set. Based on the average access time, a method is proposed to pick out the suitable solution from the set. Those suitable solutions are gathered to form the final replica location. Then we propose a method to make adjustments step by step according to the replica location. A method to reduce problem size is also proposed. MRMS is evaluated by the access history from a distributed storage cluster of Xunlei Inc. The experimental results show that MRMS can effectively improve the overall performance of the storage cluster.

Stock price in time series data can be analyzed with the clustering method by using the autocorrelation function distance measurement method. The purpose of this study is to cluster stock prices with the same characteristics and analyze companies’ financial performance in each cluster and provide a reference to investors in making choices to develop their investments. This study uses time-series data from stock prices in the LQ45 index, which is continuously available and registered from January 2010 to December 2019, as many as 32 companies. The results of this study are obtained 3 clusters, where the first cluster contains 17 stocks, the second cluster contains six stocks, and the third cluster contains nine stocks. After clustering, the financial performance of each cluster is analyzed in 2019. The companies’ financial performance in the first cluster shows that the company has proper inventories, total assets, profit for the period, and can get great benefits. The third cluster shows that the company has a relatively good current ratio and demonstrates its ability to generate profits from the high assets and equity used. Meanwhile, the second cluster has quite high receivables.

Clustering is a common phenomenon seen all around the world in industries, and the service sector. Clustering is a complicated case in retail, and mainstream literature is populated with studies that define store performance for single stores; however, not much is available when they are in clustering, as the conventional trading boundaries, which form the area in which the store's influence extends, cannot be defined. The present study was conducted to improve the overall performance of the entire cluster, by dealing with individual stores. It was conducted in a large retail cluster dealing exclusively in stationary. The store facilities are analysed using fuzzy linguistic modelling from both, the customer and the retailers stand point. A model of such clusters is then prepared for the current demographic. The model generated aims to provide a holistic approach to grade the facilities available in order to determine returns. This also gives a framework for retailers to upgrade their existing facilities according to the cluster characteristics, thus improving not only individual performance, but also the performance of the cluster.

This study aims to determine the effect of the work environment on teacher performance in Cluster VI Seririt District and to determine the effect of teacher competence on teacher performance in Cluster VI Seririt District. The type of research in this study was quantitative, with a total population of 47 teachers and a sample of 18 male teachers and 29 female teachers for a total of 47 people. The data analysis technique used in this study is multiple linear regression analysis, classical assumption test, model feasibility test (goodness of fit) using the SPSS Version 21.0 application for windows.
 The results of this study indicate that the work environment has an effect on teacher performance in Cluster VI Seririt District of 0.468. The results of testing the hypothesis with the T-test found that t-test > t-table (2.753 > 1.677). This shows that the work environment has a significant positive effect on teacher performance seen from a significant level of 0.009 <0.05. Teacher competence has an effect on teacher performance in Cluster VI Seririt District of 0.363 with the results of testing the hypothesis with the T-Test it was found that t-count > t-table (2.670 > 1.677), this shows that teacher competence has a positive and significant influence

The aims of the study are twofold: (1) to compare semantic fluency, clustering and switching performance among subjects with memory complaints, patients with Alzheimer Disease (AD), and healthy controls; and (2) to examine the clinical utility of the clustering/switching scoring system in the prediction of incident AD in subjects with memory complaints. A semantic fluency task was used to compare thirty eight subjects with memory complaints, forty two AD patients and twenty five healthy controls on the total number of words generated, clustering and switching performance. Subjects with memory complaints were followed-up for a maximum period of two years and re-evaluated. They remained in the memory complaints group (twenty eight subjects) or were defined as probable AD (ten subjects). AD patients generated fewer correct words (p < 0.001) and showed a reduction in clustering (p = 0.008) and switching (p < 0.001). Subjects with memory complaints showed a significant reduction in correct words (p < 0.001) and clustering performance (p = 0.008) compare to controls. In the first evaluation, the subgroup of patients who converted to AD at follow up produced less correct words (p < 0.01) and smaller clusters (p = 0.007) than the subgroup who did not become demented. There were no differences in switching between these two subgroups. AD development was better predicted by cluster size than by the total number of words generated or by switching. Subjects with memory complaints and AD patients have an alteration in both qualitative and quantitative aspects of semantic fluency. A clustering analysis could enhance the reliability of early AD diagnosis.

Mass cytometry is a new high-throughput technology that is becoming a cornerstone in immunology and cell biology research. With technological advancement, the number of cellular characteristics cytometry can simultaneously quantify grows, making analysis increasingly computationally onerous. In this paper, we investigate the potential of dimensionality reduction techniques to ease computational burden in clustering cytometry data whilst minimally diminishing clustering performance. We explore 3 such techniques: Principal Component Analysis (PCA), Autoencoders (AE) and Uniform Manifold Approximation and Projection (UMAP). Thereafter we employ a recent clustering algorithm, ChronoClust, which clusters data at each time-point into cell populations and explicitly tracks them over time. We evaluate this approach through a 14-dimensional cytometry dataset describing the immune response to West Nile Virus over 8 days in mice. To obtain a broad sample of clustering performance, each of the four datasets (unreduced, PCA-, AE- and UMAP-reduced) is independently clustered 400 times, using 400 unique ChronoClust parameter value sets. We find that PCA and AE can reduce the computational expense whilst incurring a minimal degradation in clustering and cluster tracking performance.