Font Size: a A A

The Application Of Concept Lattice In Data Mining

Posted on:2009-09-20Degree:MasterType:Thesis
Country:ChinaCandidate:D Y WuFull Text:PDF
GTID:2178360275461341Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of the information times, the explosive growth of data in various businesses and scientific databases has far outpaced our ability to interpret and analyze it. Data mining is the new research field against this problem. The goal is to extract implicit, previously unknown, and potentially useful information out of large amounts of collected data.Formal Concept Analysis (FCA), which was elaborated by Professor Wille of German in the eighties of the twentieth century. The Concept Lattice, which is the core data structure of FCA and is also called Galois Lattice, can describe the hierarchy relationship between concepts and has become an important method for the representation of the knowledge. As a kind of excellent mathematic tool, Concept Lattice has been widely applied in knowledge representation, data mining and many other fields.The key works in this thesis are as follows:During the process of medical image diagnosis, the method of intelligent data analysis has been widely employed in the field of medical. The using of more efficient methods to analyze the data is just what people are looking forward to. The theory of the Concept Lattice will be applied to the X-ray diagnosis of expertise database in this paper. The purpose of the intelligent analysis is achieved for the syndrome of radiograph holders through the calculation of similarity between X-ray signs.This paper presents a method of cycle association rules mining for time-series fluctuations based on the concept lattice theory. The time-series are first deseasonalized and then a new parallel algorithm of cycle association rules mining is proposed. To increase the mining speed efficiently, some concepts obtained in the algorithm are pruned. Afterward, deseasonalizeation of time-series which do not satisfy the pretreatment conditions of deseasonalizeation in moving average method is computed by a high accuracy model given in this paper. The experiment results show the validity of this method.At last, the application prospect of concept lattices and further research.
Keywords/Search Tags:data mining, Concept Lattice, intelligent data analysis, time-series, cycle association rules, moving average method, prediction model
PDF Full Text Request
Related items