search.steps

This function uses the calculated inclusion weights and selects a subset of variables in each step.
Note that it uses the values for the first target variable and first metric and might not be suitable for multi-target or multi-metric searches.

Methods and tools for model selection and multi-model inference (Burnham and Anderson (2002) <doi:10.1007/b97636>, among others).
'SUR' (for parameter estimation), 'logit'/'probit' (for binary classification), and 'VARMA' (for time-series forecasting) are implemented.
Evaluations are both in-sample and out-of-sample.
It is designed to be efficient in terms of CPU usage and memory consumption.

Ramin Mojab

Automated Uncertainty Analysis

Stephen Becker

search.steps function

<dl><dt>method</dt>
<dd>sur, bin or varma</dd>
<dt>isInnerExogenous</dt>
<dd>Determines if the inner indices are for exogenous variables.</dd>
<dt>...</dt>
<dd>Additional arguments for the search function.</dd></dl>

Arguments

Step-wise estimation — search.steps

<dl>

<dt>method</dt>
<dd>sur, bin or varma</dd>


<dt>isInnerExogenous</dt>
<dd>Determines if the inner indices are for exogenous variables.</dd>


<dt>...</dt>
<dd>Additional arguments for the search function.</dd>

</dl>

search.steps: Step-wise estimation

Description

Usage

Value

Arguments