1. Experimental Evaluation of Distance-based and Probability-based Clustering
- Author
-
Na Yeon Kwon, Richard Dollein, Jang Il Kim, Yong Gyu Jung, and Weon Joon Seo
- Subjects
Basis (linear algebra) ,business.industry ,Computer science ,Process (computing) ,k-means clustering ,Maximization ,Machine learning ,computer.software_genre ,Data set ,Data model ,Key (cryptography) ,Artificial intelligence ,Data mining ,business ,Cluster analysis ,computer - Abstract
Decision-making is to extract information that can be executed in the future, it refers to the process of discovering a new data model that is induced in the data. In other words, it is to find out the information to peel off to find the vein to catch the relationship between the hidden patterns in data. The information found here, is a process of finding the relationship between the useful patterns by applying modeling techniques and sophisticated statistical analysis of the data. It is called data mining which is a key technology for marketing database. Therefore, research for cluster analysis of the current is performed actively, which is capable of extracting information on the basis of the large data set without a clear criterion. The EM and K-means methods are used a lot in particular, how the result values of evaluating are come out in experiments, which are depending on the size of the data by the type of distance-based and probability-based data analysis. Key words: EM, Decision-making, K-Means, Maximization Step, BI-RADS assessment
- Published
- 2013
- Full Text
- View/download PDF