lpdf$optcg

A positive double representing tolerance, default is 
<code>0.001</code>.

A positive integer representing the maximum number of steps 
conjugate gradient will take.

epoch

<pre>
lpdf$optcg(tol,epoch)
</pre>This optimizes the coefficient vector <code>coeff</code> using conjugate gradient. 
It currently is designed only for quadratic <code><a rd-options="" href="/link/lpdf?package=outerbase&version=0.1.0" data-mini-rdoc="outerbase::lpdf">lpdf</a></code> instances.

High-dimensional regression using outer product models. Research on the methods is currently under investigation and published resources will be posted as they are available. As the method is new, the website is the best resource for understanding the principals. Some of the core ideas are based on Plumlee and coauthors' work on analysis of grid-structured experiments described in Plumlee (2014) <doi:10.1080/01621459.2014.900250> and Plumlee, Erickson, Ankenman, Lawrence (2021) <doi:10.1093/biomet/asaa084>. Some additional textbooks for additional information on Gaussian processes are Rasmussen and Williams (2005) <doi:10.7551/mitpress/3206.001.0001> and Gramacy (2022) <doi:10.1201/9780367815493>.

Matthew Plumlee

outerbase

Outer Product Regression

lpdf$optcg function

<pre>
lpdf$optcg(tol,epoch)
</pre>This optimizes the coefficient vector <code>coeff</code> using conjugate gradient. 
It currently is designed only for quadratic <code><a rd-options='' href='lpdf'>lpdf</a></code> instances.

lpdf$optcg: Optimization via Conjugate Gradient

Description

Arguments

Value

See Also