demos.SVR Class Reference

SVR: Support vector machine related KerMor demos. More...

Detailed Description

SVR: Support vector machine related KerMor demos.

Author

Daniel Wirtz

Date

2013-08-23

New in 0.7:

(Daniel Wirtz, 2013-08-23) Added this class.

This class is part of the framework

KerMor - Model Order Reduction using Kernels

• Homepage http://www.morepas.org/software/index.html
• Documentation http://www.morepas.org/software/kermor/index.html
• License KerMor license conditions

Definition at line 18 of file SVR.m.

Static Public Member Functions
static function PlotManager pm =	EpsLoss (double ep)
	Plots the epsilon-insensitive loss function for a given epsilon. More...
static function	ScalarEpsSVR_SMO (integer version,PlotManager pm)
	Demonstrates the general.regression.ScalarEpsSVR_SMO class. More...
static function	ScalarEpsSVR_SMO_SingleSteps (varargin)
	Demonstrates the general.regression.ScalarEpsSVR_SMO class. More...

Member Function Documentation

function PlotManager pm = demos.SVR.EpsLoss ( double  ep )

static

Plots the epsilon-insensitive loss function for a given epsilon.

Parameters

ep	The \(\epsilon\) to use. Default: 0.1

Return values

pm	A PlotManager instance to save created figures. Optional.

Generated fields of pm:

Definition at line 37 of file SVR.m.

References PlotManager.nextPlot().

Here is the call graph for this function:

function demos.SVR.ScalarEpsSVR_SMO ( integer  version, PlotManager  pm  )

static

Demonstrates the general.regression.ScalarEpsSVR_SMO class.

Call with either 1 or 2 as argument for different SMO strategies. 2 uses 2D SMO, which results in significantly faster computation time with similar results.

New in 0.5:

(Daniel Wirtz, 2011-10-05) Added this function.

This class is part of the framework

KerMor - Model Order Reduction using Kernels

• Homepage http://www.morepas.org/software/index.html
• Documentation http://www.morepas.org/software/kermor/index.html
• License KerMor license conditions

Parameters

version	The version to use. Either 1 or 2 for 1D or 2D SMO Default: 2
pm	A PlotManager instance to use for figure creation. Default: PlotManager

Required fields of pm:

Definition at line 83 of file SVR.m.

References PlotManager.LeaveOpen, and t.

function demos.SVR.ScalarEpsSVR_SMO_SingleSteps (   varargin )

static

Demonstrates the general.regression.ScalarEpsSVR_SMO class.

Call with either Version as 1 or 2 to use different SMO strategies. 2 uses 2D SMO, which results in significantly faster computation time with similar results.

New in 0.7:

(Daniel Wirtz, 2013-08-23) Added this function.

This class is part of the framework

KerMor - Model Order Reduction using Kernels

• Homepage http://www.morepas.org/software/index.html
• Documentation http://www.morepas.org/software/kermor/index.html
• License KerMor license conditions

Parameters

varargin

Various settings for the demo. ScalarEpsSVR_SMO_SingleSteps ( [ "Version", Version_value ] [, "Steps", Steps_value ] [, "PM", PM_value ] [, "Gamma", Gamma_value ] [, "Eps", Eps_value ] ) Named Parameters for varargin: Version The version to use. Either 1 or 2 for 1D or 2D SMO Default: 2 Steps The number of single iterations to perform. Default: 8 PM A PlotManager instance to use for plotting. Default: 2x2 subfigures Gamma A \(\gamma\) value for the kernel. Default: 0.8 Eps The \(\epsilon\) to use. Default: 0.1

Definition at line 143 of file SVR.m.

References l, and t.

---

The documentation for this class was generated from the following file:

• core/+demos/SVR.m