@Article{csse.2018.33.397,
AUTHOR = {Sha-Sha Li, Tie-Jun Cui, Jian Liu},
TITLE = {Research on the Clustering Analysis and Similarity in Factor Space},
JOURNAL = {Computer Systems Science and Engineering},
VOLUME = {33},
YEAR = {2018},
NUMBER = {5},
PAGES = {397--404},
URL = {http://www.techscience.com/csse/v33n5/39992},
ISSN = {},
ABSTRACT = {In this paper, we study the in uence of multiple domain attributes on the clustering analysis of object based on factor space. The representation method
of graphical domain attribute is proposed for the object, which is called attribute circle. An attribute circle can represent infinite domain attributes. The
similarity analysis of objects is first based on the concept of attribute circle, and the definition of graphical similarity is transformed into the definition of
numerical similarity, and then the clustering analysis method of object set is studied and improved. Considering three kinds of graphical overlap, the analytic
solution of similarity is obtained for numerical calculation. The clustering rules: strictly obey the similarity division and dissimilarity division, and refer to
fuzzy similarity division. The reliability evaluation semantics of the actual electrical system are listed as the study object set, and the clustering analysis
method and its improvement are carried out. The results show that the relation between decision set <i>D</i> and object set <i>U</i> means that the division of <i>U</i> is
nonsingular and accurate for <i>D</i>. Although the system reliability is evaluated in different environments, these evaluation semantics are relatively objective,
and can support each other. The two methods of similarity calculation have the same conclusion, but the improved method is more accurate and complex.},
DOI = {10.32604/csse.2018.33.397}
}