This function calculates the Mean Absolute Error (MAE) between observed and
predicted values.

An implementation of multiple regression models for
count data. These include various forms of the negative binomial (NB-1,
NB-2, NB-P, generalized negative binomial, etc.), Poisson-Lognormal, other
compound Poisson distributions, the Generalized Waring model, etc.
Information on the different forms of the negative binomial are
described by Greene (2008) <doi:10.1016/j.econlet.2007.10.015>. For
treatises on count models, see Cameron and Trivedi (2013)
<doi:10.1017/CBO9781139013567> and Hilbe (2012)
<doi:10.1017/CBO9780511973420>. For the implementation of
under-reporting in count models, see Wood et al. (2016)
<doi:10.1016/j.aap.2016.06.013>. For prediction methods in random
parameter models, see Wood and Gayah (2025)
<doi:10.1016/j.aap.2025.108147>. For estimating random parameters using
maximum simulated likelihood, see Greene and Hill (2010)
<doi:10.1108/S0731-9053(2010)26>; Gourieroux and Monfort (1996)
<doi:10.1093/0198774753.001.0001>; or Hensher et al. (2015)
<doi:10.1017/CBO9781316136232>.

Jonathan Wood

flexCountReg

Estimation of a Variety of Count Regression Models

Guillermo Basulto-Elias

mae function

<dl><dt>y</dt>
<dd>Numeric vector representing the observed values.</dd>
<dt>mu</dt>
<dd>Numeric vector representing the predicted values.</dd></dl>

Arguments

Calculate Mean Absolute Error (MAE) — mae

<dl>

<dt>y</dt>
<dd>Numeric vector representing the observed values.</dd>


<dt>mu</dt>
<dd>Numeric vector representing the predicted values.</dd>

</dl>

mae: Calculate Mean Absolute Error (MAE)

Description

Usage

Value

Arguments

Details

Examples