X.svmmaj

For efficiency use in svmmajcrossval

Implements the SVM-Maj algorithm to train data with support vector machine
<doi:10.1007/s11634-008-0020-9>.
This algorithm uses two efficient updates, one for linear kernel and one
for the nonlinear kernel.

Hok San Yip

SVMMaj

Implementation of the SVM-Maj Algorithm

Hoksan Yip

Georgi Nalbantov

X.svmmaj function

<dl><dt>object</dt>
<dd>Model which has been trained beforehand using
<code>svmmaj</code>.</dd>
<dt>X.new</dt>
<dd>Attribute matrix of the objects to be predicted, which has the
same number of attributes as the untransformed attribute matrix in
<code>model</code>.</dd>
<dt>weights</dt>
<dd>The weight of observation as the relative importance of the
prediction error of the observation.</dd></dl>

Arguments

Returns transformed attributes — X.svmmaj

<dl>

<dt>object</dt>
<dd>Model which has been trained beforehand using
<code>svmmaj</code>.</dd>


<dt>X.new</dt>
<dd>Attribute matrix of the objects to be predicted, which has the
same number of attributes as the untransformed attribute matrix in
<code>model</code>.</dd>


<dt>weights</dt>
<dd>The weight of observation as the relative importance of the
prediction error of the observation.</dd>

</dl>

X.svmmaj: Returns transformed attributes

Description

Usage

Arguments