Papers by Keyword: Cluster Analysis

Abstract: The purpose of this article is to add knowledge about the influence of preparation conditions on the resulting properties of polymeric (Nano) composites by a statistical view of the results.Most of the work on the preparation of (Nano) composites only deals with the description of the improvement or deterioration of properties without a more in-depth assessment of the statistical significance of the obtained results. In the article on the above name, the primary objective of the statistical proof of change in the strength of the tested materials will be the nanoplane content. The Yakima’s (polynomial interpolation method) will be used for the basic part of the evidence, which, by smooth interpolation, will enhance the functional dependence of the tensile strength of the modified and unmodified mixture on temperature and time.As a further one and for key evidence presented, application of cluster analysis methodology, namely the finding of a suitable calculation method based on the basic correlation matrix, will be used. The resulting evidence will then be confirmed using a further methodology, specifically cluster analysis.This article has proved to be able to prove that the obtained improvements in the properties of the prepared systems are also statistically significant.

Abstract: Article is directed to segmentation of production using hierarchical methods of cluster analysis. Analysis is performed in statistical software STATISTICA. The result of clustering process are segments of products and machinery in form of arranged machine-product matrix. These data may serve for changing the disposition of workplace toward to manufacturing cells.

Abstract: In condition monitoring (CM) signal analysis the inherent problem of key characteristics being masked by noise can be addressed by analysis of the signal envelope. Envelope analysis of vibration signals is effective in extracting useful information for diagnosing different faults. However, the number of envelope features is generally too large to be effectively incorporated in system models. In this paper a novel method of extracting the pertinent information from such signals based on multivariate statistical techniques is developed whichsubstantialy reduces the number of input parameters required for data classification models. This was achieved by clustering possible model variables into a number of homogeneous groups to assertain levels of interdependency. Representatives from each of the groups were selected for their power to discriminate between the categorical classes. The techniques established were applied to a reciprocating compressorrig wherein the target was identifying machine states with respect to operational health through comparison of signal outputs for healthy and faulty systems. The technique allowed near perfect fault classification.In addition methods for identifying seperable classes are investigated through profiling techniques, illustrated using Andrew’s Fourier curves.

Abstract: Today, load characteristic analysis plays a vital role in network planning, operation and control. In particular, with massive demand side participation activities on the network, the characteristics of individual load determines the way of the active load management as well as the network pricing strategies. In this paper, the load characteristics are analyzed utilizing clustering techniques for a tropical isle with massive temperature sensitive loads. Through deeply mining of real data, the features of individual loads were observed to define the way of load participation.

Abstract: Experimental studies very often lead to datasets with a large number of noted attributes (observed properties) and relatively small number of records (observed objects). The classic analysis cannot explain recorded attributes in the form of regression relationships due to lack of sufficient number of data points. One of method making available a filtering of unimportant attributes is an approach known as ‘dimensionality reduction’. Well-known example of such approach is principal component analysis (PCA) which transforms the data from the high-dimensional space to a space of fewer dimensions and gives heuristics to select least but necessary number of dimensions. Authors used such technique successfully in their previous investigations but a question arose: whether PCA is robust and stable This paper tries to answer this question by re-sampling experimental data and observing empirical confidence intervals of parameters used to make decision in PCA heuristics.

Abstract: The aim of this study is to discriminate herbal medicines (here after referred to as herbals) by an electronic nose (e-nose) based on an array of eight commercially gas sensors and multivariate statistical analyses. Seven kinds of herbal essential oils purchased from local market in Yogyakarta Indonesia, including zingiber officinale (ZO), kaempferia galanga (KG), curcuma longa (CL), curcuma zedoaria (CZ), languas galanga (LG), pogostemon cablin (PO), and curcuma xanthorrizha roxb (CX) were measured by using this e-nose consecutively. Due to the use of dynamic headspace in this e-nose, data for one cycle (sampling and purging) were recorded every five second for 10 cycles. Each kind of herbals was analyzed for five replications and relative amplitude of the responses was extracted as a feature. The statistical analyses of principal component analysis (PCA) and cluster analysis (CA) were used for discriminating samples. The PCA score plot shows that these 35 essential oil samples were separated into 7 groups based on similarity of patterns. The first two components, PC₁ and PC₂, capture 96.2% of data variance. Meanwhile, by using 80% similarity, the CA clusters 7 herbals into 3 classes. In this case, the first class consists of ZO and CZ and the second class consists of KG, CL, LG and CX, while the PO sample is clustered in the third class. These classes need to be validated using a standard analytical instrument such as GC/MS. The technique shows some advantages including easy in operation because of without any sample preparation, rapid detection, and good repeatability.

Abstract: The development of information technology gives rise to explosive growth of the amount of data. As a result, a more effective data mining method in pattern recognition is called into existence, which can properly reflect the inherent daily activity structure of metro travelers. This study is aimed to enrich the traditional clustering methods and provide practical information in dealing with traffic volume variation to the metro system operations. In this study, daily metro origin-destination (OD) data come from smart card records of Shenzhen, China, which cover 290 days and 118 stations. Principal component analysis (PCA) and singular value decomposition (SVD) are applied to conduct dimensionality reduction. Affinity propagation is then chosen to cluster the dimensionality reduced matrix to identify demand patterns of the metro OD matrix. Eleven representative categories are clustered and shown.

Abstract: Reservoir evaluation, which works throughout the oil and gas development, has always focused on exploration and development needs ^[1-3].There are many influencing factors and data in the evaluation process, thus it’s difficult to select the reasonable evaluation parameters and methods. Selecting the appropriate evaluation parameters and evaluation methods can improve the accuracy of evaluation, reduce the evaluation workload. Regional studies have shown that Delta-front sub-facies is the major sedimentary type of the Chang 8 in Baibao oilfield, widely distributed delta distributary channel and mouth bar sands are good oil and gas reservoirs in the area. In this paper, cluster analysis reservoir evaluation is using to obtain their correct evaluation parameters, in order to provide the basis for further development.

Abstract: This paper uses Qishan National Forest Park as a sample to apply step analysis and cluster analysis on 14 attractions in this park by GIS spatial analysis function. To be more exact, based on planar space theory, visibility, continuity, clarity, comfort and other six factors were selected to be analyzed. Results provide that one attraction has the best landscape resource, six attractions own better landscape resource, two attractions is general and five spots are poor. The results are used to verify the feasibility of landscape visual assessment model which is based on GIS technology. Furthermore, this would also provide technical support for the visual landscape assessment of forest park.

Abstract: The typical fragment screening is one of the key links in the construction driving cycle. This problem is widely adopted Duration intercept method and clustering analysis method. Both methods are subjective,and the requirements for data of working conditions is huge. Aimed at these shortages, Using multivariate statistical theory discriminate analysis method, differentiate and classify the condition data. Basing on the statistical analysis of condition data, establish the discriminate function. With a city as example, Driving Cycle is established. Based on the discriminate analysis, classify the driving data. The effectiveness of the driving cycle is confirmed. The result of comparing discriminate analysis and traditional method shows that the driving cycle based on discriminate analysis can accurately reflect the characteristics of different kinds of driving conditions, and the data demand is reduced, meanwhile, the data acquisition test is simpler, the cost of establishing a driving cycle is cut, and development cycle is reduced.