energylossbeta

This function calculates the energy loss for given tensors. The loss is calculated
as the mean of the L2 norms between <code>yt</code> and <code>mxt</code> and between <code>yt</code> and <code>mxpt</code>,
each raised to the power of <code>beta</code>, subtracted by half the mean of the L2 norm between <code>mxt</code> and <code>mxpt</code>,
also raised to the power of <code>beta</code>.

internal

Fits engression models for nonlinear distributional regression. Predictors and targets can be univariate or multivariate. Functionality includes estimation of conditional mean, estimation of conditional quantiles, or sampling from the fitted distribution. Training is done full-batch on CPU (the python version offers GPU-accelerated stochastic gradient descent). Based on "Engression: Extrapolation through the lens of distributional regression" by Xinwei Shen and Nicolai Meinshausen (2024) in JRSSB. Also supports classification (experimental).
<doi:10.1093/jrsssb/qkae108>.

Nicolai Meinshausen

engression

Engression Modelling

Xinwei Shen

energylossbeta function

<dl><dt>yt</dt>
<dd>A tensor representing the target values.</dd>
<dt>mxt</dt>
<dd>A tensor representing the model's stochastic predictions.</dd>
<dt>mxpt</dt>
<dd>A tensor representing another draw of the model's stochastic predictions.</dd>
<dt>beta</dt>
<dd>A numeric value for scaling the energy loss.</dd></dl>

Arguments

Energy Loss Calculation with Beta Scaling — energylossbeta

<dl>

<dt>yt</dt>
<dd>A tensor representing the target values.</dd>


<dt>mxt</dt>
<dd>A tensor representing the model's stochastic predictions.</dd>


<dt>mxpt</dt>
<dd>A tensor representing another draw of the model's stochastic predictions.</dd>


<dt>beta</dt>
<dd>A numeric value for scaling the energy loss.</dd>

</dl>

energylossbeta: Energy Loss Calculation with Beta Scaling

Description

Usage

Value

Arguments