rm_class-class

rm_class

A novel ensemble method employing Support Vector Machines (SVMs) as base learners. This powerful ensemble model is designed for both classification (Ara A., et. al, 2021) <doi:10.6339/21-JDS1014>, and regression (Ara A., et. al, 2021) <doi:10.1016/j.eswa.2022.117107> problems, offering versatility and robust performance across different datasets and compared with other consolidated methods as Random Forests (Maia M, et. al, 2021) <doi:10.6339/21-JDS1025>.

Mateus Maia

randomMachines

An Ensemble Modeling using Random Machines

Anderson Ara

Gabriel Ribeiro

rm_class-class function

<dl>
<dt><code>train</code></dt>
<dd>a <code>data.frame</code> corresponding to the training data used into the model</dd></dl>
<dt><code>class_name</code></dt>
<dd>a string with target variable used in the model</dd>
<dt><code>kernel_weight</code></dt>
<dd>a numeric vector corresponding to the weights for each bootstrap model contribution</dd>
<dt><code>lambda_values</code></dt>
<dd>a named list with value of the vector of \(\boldsymbol{\lambda}\) sampling probabilities associated with each each kernel function</dd>
<dt><code>model_params</code></dt>
<dd>a list with all used model specifications</dd>
<dt><code>bootstrap_models</code></dt>
<dd>a list with all <code>ksvm</code> objects for each bootstrap sample</dd>
<dt><code>bootstrap_samples</code></dt>
<dd>a list with all bootstrap samples used to train each base model of the ensemble</dd>
<dt><code>prob</code></dt>
<dd>a boolean indicating if a probabilitistic approch was used in the classification Random Machines</dd>

Slots

S4 class for RM classification — rm_class-class

<dl>
<dt><code>train</code></dt>
<dd>a <code>data.frame</code> corresponding to the training data used into the model</dd>


<dt><code>class_name</code></dt>
<dd>a string with target variable used in the model</dd>


<dt><code>kernel_weight</code></dt>
<dd>a numeric vector corresponding to the weights for each bootstrap model contribution</dd>


<dt><code>lambda_values</code></dt>
<dd>a named list with value of the vector of \(\boldsymbol{\lambda}\) sampling probabilities associated with each each kernel function</dd>


<dt><code>model_params</code></dt>
<dd>a list with all used model specifications</dd>


<dt><code>bootstrap_models</code></dt>
<dd>a list with all <code>ksvm</code> objects for each bootstrap sample</dd>


<dt><code>bootstrap_samples</code></dt>
<dd>a list with all bootstrap samples used to train each base model of the ensemble</dd>


<dt><code>prob</code></dt>
<dd>a boolean indicating if a probabilitistic approch was used in the classification Random Machines</dd>


</dl>

rm_class-class: S4 class for RM classification

Description

Arguments

Slots

Details