Papers by Keyword: Hierarchical Clustering

Abstract: The conventional framework for Structural Health Monitoring (SHM) primarily focuses on individual structures. However, to effectively identify the most vulnerable elements, preliminary studies are required at a wide area scale. This becomes particularly challenging in urban settings, where numerous buildings of varied shapes, ages, and structural conditions are closely spaced from one another. A twofold task is therefore required: the automated identification and differentiation of various structures, coupled with a ranking system based on perceived structural risk, here assumed to be linked to their deformation patterns. It integrates displacement measurements acquired through the Differential Synthetic Aperture Radar Interferometry (DInSAR) technique, specifically employing the full-resolution Small Baseline Subset (SBAS) approach, with Hierarchical Clustering. The effectiveness of this method is successfully demonstrated and validated in two selected areas of Rome, Italy, serving as case studies. The results achieved on this wide area scale monitoring can be used to select the constructions that need a more in-depth assessment.

Abstract: To investigate the brain default mode network (DMN) of healthy young people, a novel hierarchical clustering method was proposed to detect similarities of low-frequency fluctuations between any two out of 160 regions of interest (ROI) all over the brain. Feature of these ROIs were firstextractedand analyzed the feature using hierarchical clustering approach.Combining with the strongest connected network node identified by network centric criterion, the default mode network which presented the strongest connectivity in resting state was then determined. The results demonstrated that cingulate had the highest value of average degree, making it the most suspectof where the centrality indices of DMN lay.The comparative results between nodes included by DMN returned by our method and these given by Dosenbach’s research showed quite high coincidence rates,indicating the proposed method of combining complex network theory and hierarchical clustering analysis feasible method to parse brain regions.

Abstract: According to the analysis of text feature, the document with co-occurrence words expresses very stronger and more accurately topic information. So this paper puts forward a text clustering algorithm of word co-occurrence based on association-rule mining. The method uses the association-rule mining to extract those word co-occurrences of expressing the topic information in the document. According to the co-occurrence words to build the modeling and co-occurrence word similarity measure, then this paper uses the hierarchical clustering algorithm based on word co-occurrence to realize text clustering. Experimental results show the method proposed in this paper improves the efficiency and accuracy of text clustering compared with other algorithms.

Abstract: With the increasingly complex electromagnetic environment and continuous appearance of advanced system radars, signals received by radar reconnaissance receivers become even more intensive and complex, because scanning time of radar reconnaissance of each direction is very small, composition of the signals received are very complex, number of signals from different radar emitters differ greatly, traditional radar sorting methods can’t process the signals effectively. Aiming at solving the above problem, a novel radar multi-parameter signal sorting method based on data field and hierarchical clustering is proposed. Data field is introduced to reduce compute capacity and determine the parameters of hierarchical clustering. Hierarchical clustering is known that can obtain multi-level clustering structure of different particle size, by which, we can get small number of radar signals drown in many radar signals. Experimental results show that method presented in this paper can sort radar signals in complex electromagnetic environment effectively.

Abstract: Heuristic methods by first order sensitivity analysis are often used to determine location of capacitors of distribution power system. The selected nodes by first order sensitivity analysis often have virtual high by first order sensitivities, which could not obtain the optimal results. This paper presents an effective method to optimally determine the location and capacities of capacitors of distribution systems, based on an innovative approach by the second order sensitivity analysis and hierarchical clustering. The approach determines the location by the second order sensitivity analysis. Comparing with the traditional method, the new method considers the nonlinear factor of power flow equation and the impact of the latter selected compensation nodes on the previously selected compensation location. This method is tested on a 28-bus distribution system. Digital simulation results show that the reactive power optimization plan with the proposed method is more economic while maintaining the same level of effectiveness.

Abstract: This paper is based on Kansei engineering theories and methods combines with hierarchical clustering and gray correlation theory. The information of consumers perceptual demands for the product is obtained, and after that the existing products are collected and sifted, and their design elements are analyzed. The information of consumers perceptual demands is classified and extracted by using systematic clustering method perceptual demands. Two incidence matrixes have been created by investigating consumers perceptual demands. Finally, we get the incidence matrix between consumers perceptual demands and existing product samples. Also, analysis of correlation between each sample and the perceptual demands of consumers was made by using gray correlation theory.

Abstract: The aim of the system is to achieve query expansion. The method works combining clustering of hierarchical methods. Through the information proffered by a background document, doc_ID, concerning the initial query, a cluster containing doc_ID can be produced by hierarchical clustering. And the word co-occurrence information can be extracted from the candidate documents in this cluster. Compared with the content in doc_ID, the result of the experiment using the system shows an expected performance to develop expanded terms which are qualified to add to original query.

Abstract: Rich information is contributed to microblogs by millions of users all around the world. However, few work has been done on the study of microblog web page extraction so far. We proposed a unified structured information extraction method based on hierarchical clustering which is suitable for microblog web pages of any microblog websites. The experiment result on microblog web pages of some popular microblog service providers indicates the high performance of our method.

Abstract: Energy consumption structure optimum is gradually discussed in recent literatures. Based on hierarchical clustering of optimally close to content demand of data group mine and analysis, industrial sectors layout on carbon emission intensity is researched. Computed carbon emission drawn support from IPCC methodological framework, formed carbon emission intensities of emissions divided by sectors GDP respectively, and transformed calculated figures into CDF of the continuous uniform distribution to cultivate the standardized data. Resulting of the case presents that there are two categories with types of v and inversed v after mining and analyzing 37 industrial sectors data in 2006-2011. Findings are that 39% annual max paired difference of emission intensities is appeared, and the divergence of energy consumption structure is significantly obtained, which is conducive to the whole industrial distribution of low carbon policy-making.

Abstract: The paper mainly studies the available transfer capability of wind farm incorporated power system, and proposes an on-line calculating method considering many uncertain factors. First based on continuation power flow, an improved algorithm of the key constraint by linear prediction is proposed so as to obtain deterministic ATC with the expansion power flow equation. Then Monte Carlo simulation is used which takes many uncertain factors into considerations, such as wind speed, the random fault of generators and lines, the fluctuation of generators and load, etc. With hierarchical clustering algorithm, the samples with the same key constraint can be divided into one category without repeatedly predicting key constraint, so the probabilistic ATC in all kinds of conditions can be got rapidly and precisely. Finally a case study demonstrates the validity and superiority of method proposed in this paper. The research could be a valuable reference to power market operations and wind farm planning.