Parameter Sensitivity for Discriminant Analysis
Bibliographic Data
| ID | 19290811 |
|---|---|
| Authors | Richard L Tate, John Bryant (0000-0002-4743-1276), John L Bryant |
| Year | 1986 |
| Volume | 21 |
| Issue | 4 |
| Pages | 411-427 |
| Publication date | 1986-10-01 |
| Open Access | No |
| Type | ARTICLE |
| Venue | Multivariate Behavioral Research (JOURNAL) |
| Journal identifiers | ISSN: 0027-3171 • E-ISSN: 1532-7906 |
| Publisher | Informa UK Limited (PUBLISHER • GB) |
| DOI | 10.1207/s15327906mbr2104_3 |
| PMID | 26828219 |
| OpenAlex | W1990895530 |
| Language | EN |
| References cited | 23 |
The shape of the response surface associated with a discriminant analysis provides useful insight into the value of the derived optimal discriminant variates. If, for a given set of data, the surface is very peaked at the solution point for a specific variate, the variate is relatively unique in fitting the data. If, conversely, the surface is very flat, the associated optimal variate is not unique for practical purposes, with many other linear combinations providing a fit which is nearly as good as the optimal. A procedure for the determination of "indifference regions", presented and illustrated in this article, allows the assessment of the degree of flatness of the response surface for any analysis. Moreover, when the surface is found to be relatively flat, the results from this procedure can be used to identify alternate variates which are preferable to the optimal because of theoretical consistency while still providing an acceptable empirical fit
Consistency (knowledge bases) · Discriminant · Flatness (cosmology) · Geometry · Linear discriminant analysis · Point (geometry) · Random variable · Random variate · Sensitivity (control systems) · Set (abstract data type) · Statistics · Surface (topology) · Advanced Statistical Methods and Models · Artificial Intelligence · Computer Science · Engineering · Mathematics · Optimal Experimental Design Methods · Sensory Analysis and Statistical Methods
Discriminant Analysis
A Study of the Stability of Canonical Correlations and Canonical Components
A Monte Carlo Study of the Stability of Canonical Correlations, Canonical Weights and Canonical Variate-Variable Correlations
Four Methods of Analyzing Between Variation for the K-Group Manova Problem
The Study Of Effects In Manova
Generalized Discriminant Analysis
Parameter Sensitivity In Multivariate Methods
Multivariate ATI Analysis
Simultaneous test procedures and the choice of a test statistic in Manova
Simultaneous test procedures and the choice of a test statistic in Manova
Factorial-Design Multiple-Discriminant Analysis
Issues in the use and interpretation of discriminant analysis
On the sensitivity of regression and regressors
Canonical correlation analysis
| Highly cited | No |
|---|