create.linear.data

<code>[integer(1)]</code> number of rows to generate.

<code>[integer(1)]</code> number of columns to generate.

<code>[numeric(1)]</code> attenuation factor for beta coefficients.

<code>[numeric(1)]</code> size of first coefficient.

beta0

<code>[numeric(1)]</code> parameter for correlation matrix.

<code>[logical(1)]</code> whether to permute columns of <code>X</code> and coefficient
vector (<code>beta</code>).

permute

Create linear model data <code>Y = X * beta + epsilon</code> with
X as a <code>n * p</code> matrix of multivariate normal distributed
rows with covariance matrix<pre><code>1 rho rho^2 rho^3 ... rho^p
rho 1 rho rho^2 ... rho^(p-1)
rho^2 rho 1 rho ... rho^(p-2)
...
rho^p ...
</code></pre><code>epsilon</code> is standard normally distributed and
<code>beta[i] = beta0 * q ^ (i - 1)</code> for <code>i = 1,..., p</code>.
If <code>permute == TRUE</code>, columns of <code>X</code> as well as <code>beta</code> are permuted before
the linear model equation is evaluated to generate <code>Y</code>. These permuted
values are also the ones returned in the result.
<code>$orig.features</code> are the features with <code>beta &gt; 1 / sqrt(n)</code>.

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

create.linear.data function

<dl><dt>n</dt>
<dd><code>[integer(1)]</code> number of rows to generate.</dd>
<dt>p</dt>
<dd><code>[integer(1)]</code> number of columns to generate.</dd>
<dt>q</dt>
<dd><code>[numeric(1)]</code> attenuation factor for beta coefficients.</dd>
<dt>beta0</dt>
<dd><code>[numeric(1)]</code> size of first coefficient.</dd>
<dt>rho</dt>
<dd><code>[numeric(1)]</code> parameter for correlation matrix.</dd>
<dt>permute</dt>
<dd><code>[logical(1)]</code> whether to permute columns of <code>X</code> and coefficient
vector (<code>beta</code>).</dd></dl>

Arguments

Create linear model data <code>Y = X * beta + epsilon</code> with
X as a <code>n * p</code> matrix of multivariate normal distributed
rows with covariance matrix<pre><code>1 rho rho^2 rho^3 ... rho^p
rho 1 rho rho^2 ... rho^(p-1)
rho^2 rho 1 rho ... rho^(p-2)
...
rho^p ...
</code></pre>

<code>epsilon</code> is standard normally distributed and
<code>beta[i] = beta0 * q ^ (i - 1)</code> for <code>i = 1,..., p</code>.
If <code>permute == TRUE</code>, columns of <code>X</code> as well as <code>beta</code> are permuted before
the linear model equation is evaluated to generate <code>Y</code>. These permuted
values are also the ones returned in the result.
<code>$orig.features</code> are the features with <code>beta &gt; 1 / sqrt(n)</code>.

Create Linear Model Data — create.linear.data

<dl>

<dt>n</dt>
<dd><code>[integer(1)]</code> number of rows to generate.</dd>


<dt>p</dt>
<dd><code>[integer(1)]</code> number of columns to generate.</dd>


<dt>q</dt>
<dd><code>[numeric(1)]</code> attenuation factor for beta coefficients.</dd>


<dt>beta0</dt>
<dd><code>[numeric(1)]</code> size of first coefficient.</dd>


<dt>rho</dt>
<dd><code>[numeric(1)]</code> parameter for correlation matrix.</dd>


<dt>permute</dt>
<dd><code>[logical(1)]</code> whether to permute columns of <code>X</code> and coefficient
vector (<code>beta</code>).</dd>

</dl>

create.linear.data: Create Linear Model Data

Description

Usage

Value

Arguments

See Also