Description
Book SynopsisStructural Sensitivity in Econometric Models Edwin Kuh, John W. Neese and Peter Hollinger Provides a pathbreaking assessment of the worth of linear dynamic systems methods for probing the behavior of complex macroeconomic models. Representing a major improvement upon the standard black box approach to analyzing economic model structure, it introduces the powerful concept of parameter sensitivity analysis within a linear systems root/vector framework. The approach is illustrated with a good mediumsize econometric model (Michigan Quarterly Econometric Model of the United States). EISPACK, the Fortran code for computing characteristic roots and vectors has been upgraded and augmented by a model linearization code and a broader algorithmic framework. Also features an interface between the algorithmic code and the interactive modeling system (TROLL), making an unusually wide range of linear systems methods accessible to economists, operations researchers, engineers and physical scientists.
Table of ContentsSelected Aspects of Multivariate Analysis.
Principal Components Analysis.
Factor Analysis.
Multidimensional Scaling.
Cluster Analysis.
Multiple Regression.
Some Practical Considerations: Data Analysis Problems.
Cross-Classified Frequency Data.
Canonical Correlation Analysis.
Discriminant Analysis: The Two-Group Problem.
Multiple Discriminant Analysis and Related Topics.
Linear Structural Relations (LISREL).
Latent Structure Analysis.
Appendixes.
References.
Index.
