PreProCombClass-class


This class implements the separation of data used for
analysis and analysis of the data. The latter can include
computation of association rules as in showrules().


Preprocessing is often the most time-consuming phase in data analysis
and preprocessing transformations interdependent in unexpected
ways. This package helps to make preprocessing faster and more effective. It
provides an S4 framework for creating and evaluating preprocessing combinations
for classification, clustering and outlier detection. The framework supports
adding of user-defined preprocessors and preprocessing phases. Default preprocessors
can be used for low variance removal, missing value imputation, scaling, outlier
removal, noise smoothing, feature selection and class imbalance correction.

Markus Vattulainen

preprocomb

Tools for Preprocessing Combinations

PreProCombClass-class function

 <dl>
<dt><code>rawall</code></dt><dd>(data frame) all results</dd> <dt><code>catclassification</code></dt><dd>(data frame) classification accuracy categorized, "high" is more than 80 percent quantile value</dd> <dt><code>allclassification</code></dt><dd>(data frame) classification accuracy, means and standard deviations</dd> <dt><code>bestclassification</code></dt><dd>(data frame) best classification accuracy combinations</dd> <dt><code>allclustering</code></dt><dd>(data frame) hopkins statistics values</dd> <dt><code>bestclustering</code></dt><dd>(data frame) best hopkins statistics combinations</dd> <dt><code>alloutliers</code></dt><dd>(data frame) ORH outlier score for 95 percent quantile value</dd> <dt><code>walltime</code></dt><dd>(integer) execution time in minutes by wall time (not computation time)</dd> </dl>

PreProCombClass-class: container for combination evaluation

Description

Arguments

Slots