LibCapy - supportVectorMachine

Support vector machine class.


SVMKernel parent class




SVMKernel structure

Struct CapySVMKernelLinear :

Struct CapySVMKernelLinear's properties:

Inherits CapySVMKernel

Struct CapySVMKernelLinear's methods:


Struct CapySVMKernelPolynomial :

Struct CapySVMKernelPolynomial's properties:

Inherits CapySVMKernel

Power of the polynomial (default: 1.0)

Threshold (default: 0.0)

Struct CapySVMKernelPolynomial's methods:


Struct CapySVMKernelGaussian :

Struct CapySVMKernelGaussian's properties:

Inherits CapySVMKernel

Gamma (default: 1.0)

Struct CapySVMKernelGaussian's methods:


Struct CapySVMEvaluation :

Struct CapySVMEvaluation's properties:

Parent class

Reduction coefficient of the support vector (=1-nbSupport/nbExample)

Struct CapySVMEvaluation's methods:


Struct CapySupportVectorMachine :

Struct CapySupportVectorMachine's properties:

Inherit CapyPredictor

Reference to the used kernel

Lagrangian multipliers


Support vectors (one vector per row, input values (normalised in [0,1]) followed by category in {-1,1}

Tolerance (default: 1.0e-3)

Margin relaxation coefficient (default: 1e-2, the higher the more sensitive to outliers, must be strictly greater than that.tolerance)

Seed for the pseudo random generator used to shuffle the rows during training

Index of the predicted category (default: 1)

Reduction coefficient of the support vector (=1-nbSupport/nbExample), used during evaluation of the SVM

Maximum number of iteration during training (default: 0, if equal to 0 the number rows in the dataset is used instead)

Flag for verbose mode (default: false)

Struct CapySupportVectorMachine's methods:


Function herited from the parent to export the body to HTML.

Input argument(s):

stream: the stream where to export
title: the title of the web app
dataset: the training dataset
expectedAccuracy: the expected accuracy of the predictor (in [0,1])

Output and side effect(s):

The and part of a ready to use web app implementing the predictor is written on the stream. The web app can be completed by calling exportToHtml on the predictor to write the