The Exponential kernel class.
Note that it can handle a length scale for each dimension for
Automtic Relevance Determination.
|
__init__(self,
length_scale=1.0,
sigma_f=1.0,
**kwargs)
Initialize an Exponential kernel instance. |
source code
|
|
|
|
|
|
|
gradient(self,
data1,
data2)
Compute gradient of the kernel matrix. A must for fast
model selection with high-dimensional data. |
source code
|
|
|
|
|
compute_lml_gradient(self,
alphaalphaT_Kinv,
data)
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. |
source code
|
|
|
compute_lml_gradient_logscale(self,
alphaalphaT_Kinv,
data)
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. |
source code
|
|
Inherited from Kernel :
compute_gradient ,
reset
Inherited from object :
__delattr__ ,
__format__ ,
__getattribute__ ,
__hash__ ,
__new__ ,
__reduce__ ,
__reduce_ex__ ,
__setattr__ ,
__sizeof__ ,
__str__ ,
__subclasshook__
|