| Summary: | Flexible discriminant analysis is development of linear discriminant
analysis, which is used to solve the problem of classification on non-parametric
regression. Flexible discriminant analysis replaces parametric regression with
non-parametric regression, one of the non-parametric regression can be used is
Multivariate Adaptive Regression Splines (MARS). MARS aims to predict, the
best MARS models which is have minimum of Generalized Cross Validation
(GCV). The assumptions is used in flexible discriminant analysis modeling is that
there is no outlier data, homogeneity variance-covariance matrix. In case study,
classify dissolved oxygen levels of water based on 5 predictor variables.
|
|---|