A comparison of two approaches to fitting directed graphs to nonsymmetric proximity measures |
| |
| Authors: | K C Klauer J D Carroll |
| |
| Institution: | 1. Institut für Psychologie, Freie Universit?t Berlin, Habelschwerdter Allee 45, 1000, Berlin 33, FR Germany
2. Graduate School of Management, Rutgers University, 92 New Street, 07102-1895, Newark, NJ, USA
|
| |
| Abstract: | Two algorithms for fitting directed graphs to nonsymmetric proximity data are compared. The first approach, termed MAPNET,
is a direct extension of a mathematical programming procedure for fitting undirected graphs to symmetric proximity data presented
by Klauer and Carroll (1989). For a user-specified number of links, the algorithm seeks to provide the connected network that
gives the least-squares approximation of the proximity data with the specified number of links, allowing for linear transformations
of the data. The mathematical programming approach is compared to the NETSCAL method for fitting directed graphs (Hutchinson
1989), using the Monte Carlo methods and data sets employed by Hutchinson. |
| |
| Keywords: | Multivariate data analysis ordinal network representation NETSCAL proximities |
| 本文献已被 SpringerLink 等数据库收录！ |
|
