bootstrap_se

the fitted values from a flexible estimation technique
regressing Y on X.

the fitted values from a flexible estimation technique
regressing either (a) <code>f1</code> or (b) Y on X withholding the columns in
<code>indx</code>.

the type of importance to compute; defaults to
<code>r_squared</code>, but other supported options are <code>auc</code>,
<code>accuracy</code>, <code>deviance</code>, and <code>anova</code>.

type

the number of bootstrap replicates (defaults to 1000)

Compute bootstrap-based standard error estimates for variable importance

Calculate point estimates of and valid confidence intervals for
nonparametric, algorithm-agnostic variable importance measures in high and low dimensions,
using flexible estimators of the underlying regression functions. For more information
about the methods, please see Williamson et al. (Biometrics, 2020), Williamson et al. (arXiv, 2020+) <arXiv:2004.03683>, and Williamson and Feng (ICML, 2020).

Brian D. Williamson

vimp

Perform Inference on Algorithm-Agnostic Variable Importance

Jean Feng

Noah Simon

Marco Carone

bootstrap_se function

<dl><dt>Y</dt>
<dd>the outcome.</dd>
<dt>f1</dt>
<dd>the fitted values from a flexible estimation technique
regressing Y on X.</dd>
<dt>f2</dt>
<dd>the fitted values from a flexible estimation technique
regressing either (a) <code>f1</code> or (b) Y on X withholding the columns in
<code>indx</code>.</dd>
<dt>type</dt>
<dd>the type of importance to compute; defaults to
<code>r_squared</code>, but other supported options are <code>auc</code>,
<code>accuracy</code>, <code>deviance</code>, and <code>anova</code>.</dd>
<dt>b</dt>
<dd>the number of bootstrap replicates (defaults to 1000)</dd></dl>

Arguments

Compute bootstrap-based standard error estimates for variable importance — bootstrap_se

<dl>

<dt>Y</dt>
<dd>the outcome.</dd>


<dt>f1</dt>
<dd>the fitted values from a flexible estimation technique
regressing Y on X.</dd>


<dt>f2</dt>
<dd>the fitted values from a flexible estimation technique
regressing either (a) <code>f1</code> or (b) Y on X withholding the columns in
<code>indx</code>.</dd>


<dt>type</dt>
<dd>the type of importance to compute; defaults to
<code>r_squared</code>, but other supported options are <code>auc</code>,
<code>accuracy</code>, <code>deviance</code>, and <code>anova</code>.</dd>


<dt>b</dt>
<dd>the number of bootstrap replicates (defaults to 1000)</dd>

</dl>

bootstrap_se: Compute bootstrap-based standard error estimates for variable importance

Description

Usage

Value

Arguments