The Exponential kernel class.
Note that it can handle a length scale for each dimension for Automtic Relevance Determination.
Initialize instance of ExponentialKernel
Parameters : | length_scale :
sigma_f :
enable_ca : None or list of str
disable_ca : None or list of str
|
---|
Compute grandient of the kernel and return the portion of log marginal likelihood gradient due to the kernel. Shorter formula. Allows vector of lengthscales (ARD) BUT THIS LAST OPTION SEEMS NOT TO WORK FOR (CURRENTLY) UNKNOWN REASONS.
Compute grandient of the kernel and return the portion of log marginal likelihood gradient due to the kernel. Shorter formula. Allows vector of lengthscales (ARD). BUT THIS LAST OPTION SEEMS NOT TO WORK FOR (CURRENTLY) UNKNOWN REASONS.
Compute gradient of the kernel matrix. A must for fast model selection with high-dimensional data.