Length scale in Kernel functions

Hello all,

I am new to machine learning and I have a question regarding ARD Matern 5/2 kernel functions used in Bayesopt.

How is characteristic length scale calculated in? Is it a constnat value or it varies in each iteration? Any useful refenrece?

The documentation is not very clear to me and any helps is greatly appreciated.

Answer 1

[Anshika Chaurasia]

Hi,

"The characteristic length scales briefly define how far apart the input values can be for the response values to become uncorrelated. "

Refer to the following documentation for more information on characteristic length scale:

Kernel (Covariance) Function Options - MATLAB & Simulink - MathWorks India

Hope it helps!