|
Param
A class for training a classifier with several values of a parameter.
|
|
ParamGrid
A class for training and testing a classifier on a grid of parameter
values for two attributes of the classifier.
|
|
ModelSelector
A model selector decides on the best classifier parameters
using the param object it receives as input.
|
|
SVMselect
A model selector for searching for best parameters for an
SVM classifier with a Gaussian kernel
Its search strategy is as follows:
First optimize the width of the Gaussian (gamma) for a fixed (low)
value of C, and then optimize C.
|