Gaussian process regression Beta values

This is how I use my fitgrp to train my data...

regressionGP_trial = fitrgp(predictors,response, 'BasisFunction', 'pureQuadratic', 'KernelFunction','exponential', 'FitMethod','sd', 'HyperparameterOptimizationOptions',struct('UseParallel','true'), 'Standardize', true);

After executing my Beta values are as follows: 0.250263965622353, 0.0563818610636318, 0.0558561922570860, -0.0177372372437889, -0.0532134171662675, 0.0133177038469632, -0.000772965282431688

What are these Beta values before executing the above function and how is it calculated?

Please advice

Answer 1

[Gautam Pendse]

Hi Sareena,

The initial value for 'Beta' is a vector of all zeros. You can specify an initial value using the 'Beta' name/value pair:

https://www.mathworks.com/help/stats/fitrgp.html#namevaluepairarguments

Hope this helps,

Gautam