Clustering Algorithm Based On Lattice Theory And Its Application In Early Colorectal Cancer Diagnosis System Research

Data mining is one of the activest branches in fields of research, development and application on database. It gets much attention since being put forward. We can discover a lot of implied, valuable and potential knowledge from mass data through data mining.As a main task of data mining,cluster analysis has been researched for many years and a lot of related theories, methods have been accumulated. In pace with development of database technology, information data from various walks of life increase rapidly. Moreover, the data types chang from pure numerical or nominal to mixed, that cause much difficulty for cluster analysis. Most algorithms can work well when the data should be processed is single type, but the performance are very low if the data is mixed type. In this paper, we put emphasis on clustering methods for mixed data.After studing the exsiting clusterting algorithms, we propose a new method that based on lattic. We call . In this algorithm,data space should be divided into many Lattices, the similarity of two objects could be described by the count of lattices that cover both of them. According to this principle, a object could be dispatch to the nearest cluster. The experimental result shows, that new algorithm can promote the precision of clustering, especially in mixed data.According with the new algorithm, optimizing parameters are put forward: 1. A method based on genetic algorithm for optimizing number of clusters. The number of cluster product by this method is very close to the real number of clusters. 2. A new method for optimizing initial cluster centers. There is further promotion of clustering precision after optimizing initial cluster center with this new method.Finally, a complete Colorectal Carcinoma Diagnostic data processing System based on data mining is designed and the cluster analysis subsystem has been implemented.