SVMUtilities

Full name Microsoft.VisualBasic.MachineLearning.SVM.SVMUtilities Assembly Microsoft.VisualBasic.MachineLearning Members 7

svm demo test

01 Syntax

Microsoft.VisualBasic.MachineLearning.SVM.SVMUtilities

02 Methods

Name	Overloads	Summary
CreateTwoClassProblem	1	Create a randomized two class classification problem for the demo test.
CreateMulticlassProblem	1	Create a randomized multiclass classification problem for the demo test.
CreateRegressionProblem	1	Create a randomized support vector regression problem for the demo test: the label value is calculated as x = 2 * y + z.

03 Fields

Name	Overloads	Summary
TRAINING_SEED	2	The random seed value which is used for generating the training data.
TESTING_SEED	2	The random seed value which is used for generating the testing data.

04 Members

method CreateTwoClassProblem #

CreateTwoClassProblem(Int32, Boolean)

Create a randomized two class classification problem for the demo test.

Parameters

Name	Type	Description
`count`	`Int32`	The total number of the samples.
`isTraining`	`Boolean`	`True` to use the SVMUtilities.TRAINING_SEED random seed, `False` to use the SVMUtilities.TESTING_SEED random seed.

Returns

A Problem object in which the points are located on either side of the vertical axis.

method CreateMulticlassProblem #

CreateMulticlassProblem(Int32, Int32, Boolean)

Create a randomized multiclass classification problem for the demo test. The points of each class are located in one of the eight octants of the three dimensional space.

Parameters

Name	Type	Description
`numberOfClasses`	`Int32`	The number of the target classes, this value should be less than 8.
`count`	`Int32`	The total number of the samples.
`isTraining`	`Boolean`	`True` to use the SVMUtilities.TRAINING_SEED random seed, `False` to use the SVMUtilities.TESTING_SEED random seed.

Returns

A Problem object which contains the randomized samples.

method CreateRegressionProblem #

CreateRegressionProblem(Int32, Boolean)

Create a randomized support vector regression problem for the demo test: the label value is calculated as x = 2 * y + z.

Parameters

Name	Type	Description
`count`	`Int32`	The total number of the samples.
`isTraining`	`Boolean`	`True` to use the SVMUtilities.TRAINING_SEED random seed, `False` to use the SVMUtilities.TESTING_SEED random seed.

Returns

A Problem object which contains the randomized samples.

field TRAINING_SEED #

TRAINING_SEED

The random seed value which is used for generating the training data.

field TESTING_SEED #

TESTING_SEED

The random seed value which is used for generating the testing data.

field TRAINING_SEED overload 2 #

TRAINING_SEED

field TESTING_SEED overload 2 #

TESTING_SEED