Tool NameSVM Classification
Tool ID0
Library IDimagery_svm
Version1.0
Author(s)O.Conrad (c) 2012

Description

Support Vector Machine (SVM) based classification for grids.

References

Chang, C.-C. & Lin, C.-J. (2011): A library for support vector machines. ACM Transactions on Intelligent Systems and Technology, vol.2/3, p.1-27. LIBSVM Homepage


Parameters

NameTypeIdentifierDescriptionConstraints
Input
Gridsgrid list, inputGRIDS
Training Areasshapes, inputROI
Output
Classificationgrid, outputCLASSES
Look-up Tabletable, output, optionalCLASSES_LUTA reference list of the grid values that have been assigned to the training classes.
Options
ScalingchoiceSCALINGAvailable Choices:
[0] none
[1] normalize (0-1)
[2] standardize
Default: 2
Verbose MessagesbooleanMESSAGEDefault: 0
Model SourcechoiceMODEL_SRCAvailable Choices:
[0] create from training areas
[1] restore from file
Default: 0
Restore Model from Filefile pathMODEL_LOAD
Class Identifiertable fieldROI_ID
Store Model to Filefile pathMODEL_SAVE
SVM TypechoiceSVM_TYPEAvailable Choices:
[0] C-SVC
[1] nu-SVC
[2] one-class SVM
[3] epsilon-SVR
[4] nu-SVR
Default: 0
Kernel TypechoiceKERNEL_TYPElinear: u'*v
polynomial: (gamma*u'*v + coef0)^degree
radial basis function: exp(-gamma*|u-v|^2)
sigmoid: tanh(gamma*u'*v + coef0)
Available Choices:
[0] linear
[1] polynomial
[2] radial basis function
[3] sigmoid
Default: 2
Degreeinteger numberDEGREEdegree in kernel functionDefault: 3
Gammafloating point numberGAMMAgamma in kernel functionDefault: 0.000000
coef0floating point numberCOEF0coef0 in kernel functionDefault: 0.000000
Cfloating point numberCOSTparameter C (cost) of C-SVC, epsilon-SVR, and nu-SVRDefault: 1.000000
nu-SVRfloating point numberNUparameter nu of nu-SVC, one-class SVM, and nu-SVRDefault: 0.500000
SVR Epsilonfloating point numberEPS_SVRepsilon in loss function of epsilon-SVRDefault: 0.100000
Cache Sizefloating point numberCACHE_SIZEcache memory size in MBDefault: 100.000000
Epsilonfloating point numberEPStolerance of termination criterionDefault: 0.001000
ShrinkingbooleanSHRINKINGwhether to use the shrinking heuristicsDefault: 0
Probability EstimatesbooleanPROBABILITYwhether to train a SVC or SVR model for probability estimatesDefault: 0
Cross Validationinteger numberCROSSVALn-fold cross validation: n must > 1Minimum: 1
Default: 1