A GRAPH THEORETICAL APPROACH TO CLUSTERING
KANSAS UNIV LAWRENCE DEPT OF PSYCHOLOGY
Pagination or Media Count:
The report describes a method of clustering variables based upon graph theory. The clusters computed are the strong and weak components of a digraph. This program was developed because of the absence of suitable programs for clustering variables when all of the relationships between the variables were not symmetric. The major part of the report describes the program, the computer system of which it is a part, and gives an example as to how it works together with sample output from the program.
- Statistics and Probability