unbiasedHoldoutDomHV

<code>[matrix]</code> fitness matrix of training data.

fitness

<code>[matrix]</code> fitness matrix of holdout data.

holdout

<code>[numeric]</code> reference point.

refpoint

Calculate dominated hypervolume on holdout data. The result is
unbiased with respect to (uncorrelated with respect to objectives) noise in holdout data
performance, but it is not an estimate of real "dominated hypervolume".
Only works on two-objective performance matrices.

Performs simultaneous hyperparameter tuning and
feature selection through both single-objective and multi-objective
optimization as described in Binder, Moosbauer et al. (2019) <arXiv:1912.12912>.
Uses the 'ecr'-package as basis but adds mixed integer
evolutionary strategies and multi-fidelity functionality as well as operators
specific for the problem of feature selection.

Martin Binder

mosmafs

Multi-Objective Simultaneous Model and Feature Selection

Susanne Dandl

Julia Moosbauer

unbiasedHoldoutDomHV function

<dl><dt>fitness</dt>
<dd><code>[matrix]</code> fitness matrix of training data.</dd>
<dt>holdout</dt>
<dd><code>[matrix]</code> fitness matrix of holdout data.</dd>
<dt>refpoint</dt>
<dd><code>[numeric]</code> reference point.</dd></dl>

Arguments

Unbiased Dominated Hypervolume on Holdout Data — unbiasedHoldoutDomHV

<dl>

<dt>fitness</dt>
<dd><code>[matrix]</code> fitness matrix of training data.</dd>


<dt>holdout</dt>
<dd><code>[matrix]</code> fitness matrix of holdout data.</dd>


<dt>refpoint</dt>
<dd><code>[numeric]</code> reference point.</dd>

</dl>

unbiasedHoldoutDomHV: Unbiased Dominated Hypervolume on Holdout Data

Description

Usage

Value

Arguments