vglm.fit.coef

Estimate the coefficients of a vector generalized linear model via parallel
iterative re-weighted least squares. Computations can be performed in parallel
to speed up the execution.

internal

Efficient framework to estimate high-dimensional generalized matrix factorization models using penalized maximum likelihood under a dispersion exponential family specification. Either deterministic and stochastic methods are implemented for the numerical maximization. In particular, the package implements the stochastic gradient descent algorithm with a block-wise mini-batch strategy to speed up the computations and an efficient adaptive learning rate schedule to stabilize the convergence. All the theoretical details can be found in Castiglione et al. (2024, <doi:10.48550/arXiv.2412.20509>). Other methods considered for the optimization are the alternated iterative re-weighted least squares and the quasi-Newton method with diagonal approximation of the Fisher information matrix discussed in Kidzinski et al. (2022, <http://jmlr.org/papers/v23/20-1104.html>).

Cristian Castiglione

sgdGMF

Estimation of Generalized Matrix Factorization Models via
Stochastic Gradient Descent

Davide Risso

Alexandre Segers

vglm.fit.coef function

<dl><dt>Y</dt>
<dd>\(n \times m\) matrix of response variables</dd>
<dt>X</dt>
<dd>\(n \times p\) matrix of covariates</dd>
<dt>family</dt>
<dd>a <code>glm</code> family (see <code><a href="/link/family?package=sgdGMF&version=1.0" data-mini-rdoc="sgdGMF::family">family</a></code> for more details)</dd>
<dt>weights</dt>
<dd>\(n \times m\) matrix of weighting values</dd>
<dt>offset</dt>
<dd>\(n \times m\) matrix of offset values</dd>
<dt>parallel</dt>
<dd>if <code>TRUE</code>, allows for parallel computing using the <code>foreach</code> package</dd>
<dt>nthreads</dt>
<dd>number of cores to be used in parallel (only if <code>parallel=TRUE</code>)</dd>
<dt>clust</dt>
<dd>registered cluster to be used for distributing the computations (only if <code>parallel=TRUE</code>)</dd></dl>

Arguments

Estimate the coefficients of a vector generalized linear model — vglm.fit.coef

<dl>

<dt>Y</dt>
<dd>\(n \times m\) matrix of response variables</dd>


<dt>X</dt>
<dd>\(n \times p\) matrix of covariates</dd>


<dt>family</dt>
<dd>a <code>glm</code> family (see <code><a href='https://rdrr.io/r/stats/family.html'>family</a></code> for more details)</dd>


<dt>weights</dt>
<dd>\(n \times m\) matrix of weighting values</dd>


<dt>offset</dt>
<dd>\(n \times m\) matrix of offset values</dd>


<dt>parallel</dt>
<dd>if <code>TRUE</code>, allows for parallel computing using the <code>foreach</code> package</dd>


<dt>nthreads</dt>
<dd>number of cores to be used in parallel (only if <code>parallel=TRUE</code>)</dd>


<dt>clust</dt>
<dd>registered cluster to be used for distributing the computations (only if <code>parallel=TRUE</code>)</dd>

</dl>

Last chance! 50% off unlimited learning

vglm.fit.coef: Estimate the coefficients of a vector generalized linear model

Description

Usage

Arguments