Noone?
Matlab GPR -> equation for prediction
11 views (last 30 days)
Show older comments
Dear MathWorks-Team
I have a question regarding a GPR model (classreg.learning.regr.CompactRegressionGP). I trained a model in Matlab which works very well. The problem is, I want to use it in javascript for a webpage and this is where my issue starts. I know, that there is possibility to export a model using Matlab Coder into c++/c this does not help in my case since I really need to implement it myself in the context of this webpage.
My knowledge on GP is however only basic and I struggle to understand the theory here (https://ch.mathworks.com/help/stats/exact-gpr-method.html). I know that I can derive the hyperparameters, ect from here (https://ch.mathworks.com/help/stats/classreg.learning.regr.compactregressiongp-class.html). The issue is, this is for someone who is not very very deep in the material impossible to understand/implent himself.
Could you therefore please write me down an easy equation based on the parameters of a model, how I can rewrite the prediction function?
This is my model:
obj_TrainedModel =
classreg.learning.regr.CompactRegressionGP
PredictorNames: {'column_2' 'column_3' 'column_4'}
ResponseName: 'Y'
CategoricalPredictors: []
ResponseTransform: 'none'
KernelFunction: 'Matern52'
KernelInformation: [1×1 struct]
BasisFunction: 'Constant'
Beta: 0.5967
Sigma: 0.2731
PredictorLocation: [3×1 double]
PredictorScale: [3×1 double]
Alpha: [39×1 double]
ActiveSetVectors: [39×3 double]
PredictMethod: 'Exact'
ActiveSetSize: 39
I appreciate all help a lot! Thanks in advance,
Best regards, J
Accepted Answer
More Answers (0)
See Also
Categories
Find more on Logical in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!