Jan Hendrik Metzen
|
3957fc3b34
|
ENH GPR.log_marginal_likelihood() returns the current log-likelihood if no theta vector is provided
|
2015-10-19 18:07:14 +02:00 |
|
Jan Hendrik Metzen
|
e2a340a4ee
|
REF More meaningful names for hyperparameters of GP kernels
|
2015-10-19 18:07:14 +02:00 |
|
Jan Hendrik Metzen
|
65555b300b
|
REF Renaming sigma_squared_n to alpha (consistency with Ridge)
|
2015-10-19 18:07:13 +02:00 |
|
Jan Hendrik Metzen
|
c6db24e56f
|
REF Several small changes based on @eickenberg's suggestions
|
2015-10-19 18:07:13 +02:00 |
|
Jan Hendrik Metzen
|
588cabfc46
|
TST Added test checking that GPR can deal with different outputs for same input
|
2015-10-19 18:07:12 +02:00 |
|
Jan Hendrik Metzen
|
2839f66933
|
REFACTOR normalize_y in GPR does not modify standard-deviation
|
2015-10-19 18:07:12 +02:00 |
|
Jan Hendrik Metzen
|
bcd4b72897
|
REFACTOR Remove theta_ attribute from GPR and GPC
It can be accessed instead via kernel_.theta
|
2015-10-19 18:07:11 +02:00 |
|
Jan Hendrik Metzen
|
15309a4227
|
TST Adapting GP tests to changed hyperparameter representation
|
2015-10-19 18:07:11 +02:00 |
|
Jan Hendrik Metzen
|
017a7fbe13
|
TST Testing GPR with custom external optimizer
|
2015-10-19 18:07:11 +02:00 |
|
Jan Hendrik Metzen
|
e813a73e3d
|
TST Testing multi-out GPR
|
2015-10-19 18:07:11 +02:00 |
|
Jan Hendrik Metzen
|
d7784969b4
|
TST Testing normalization of output values in GPR
|
2015-10-19 18:07:11 +02:00 |
|
Jan Hendrik Metzen
|
95d6577427
|
TST An increasing number of random-starts of GPR only increases the LML of the chosen theta.
|
2015-10-19 18:07:10 +02:00 |
|
Jan Hendrik Metzen
|
ebf3f05994
|
REFACTOR Removed literal form of specifying bounds for ConstantKernel
|
2015-10-19 18:07:10 +02:00 |
|
Jan Hendrik Metzen
|
31ef64571e
|
FIX Dealing with anisotropic length-scales correctly
|
2015-10-19 18:07:10 +02:00 |
|
Jan Hendrik Metzen
|
4571a7f648
|
TST Adapted tests to changes in kernels and added additional test for theta property
|
2015-10-19 18:07:10 +02:00 |
|
Jan Hendrik Metzen
|
0a23d9e604
|
REFACTOR Compute standard deviation of predictive distribution at n points in O(n)
|
2015-10-19 18:04:56 +02:00 |
|
Jan Hendrik Metzen
|
141eb0b27c
|
REFACTOR Rename GPR method sample to sample_y
|
2015-10-19 18:04:56 +02:00 |
|
Jan Hendrik Metzen
|
e3423b989a
|
FIX Minor bugfixes and downward-compatibility/python3 issue
|
2015-10-19 18:04:56 +02:00 |
|
Jan Hendrik Metzen
|
403b5f1412
|
REFACTOR REnamed kernel property params to theta
|
2015-10-19 18:04:55 +02:00 |
|
Jan Hendrik Metzen
|
9f6e48f6cc
|
REFACTOR GP kernels use separate specification of bounds and params (no joint param_space)
|
2015-10-19 18:04:55 +02:00 |
|
Jan Hendrik Metzen
|
d0bc5dc35c
|
FIX GPs do not modify kernel attribute but store optimized kernels in kernel_
|
2015-10-19 18:04:55 +02:00 |
|
Jan Hendrik Metzen
|
555bb11554
|
FIX test_converged_to_local_maximum deals with convergence to boundary
|
2015-10-19 18:04:55 +02:00 |
|
Jan Hendrik Metzen
|
cb7c4243e8
|
MISC Reorganizing kernel module
|
2015-10-19 18:04:55 +02:00 |
|
Jan Hendrik Metzen
|
d8405c2403
|
REFACTOR Hyperparameter optimization in GPs enabled explicitly by optimizer parameter
|
2015-10-19 18:04:55 +02:00 |
|
Jan Hendrik Metzen
|
e165fed1cc
|
REFACTOR GaussianProcessClassification/Regression -> ...Classifier/Regressor
|
2015-10-19 18:04:55 +02:00 |
|
Jan Hendrik Metzen
|
a5dc3491ad
|
MISC Revised gaussian_process package based on @eickenberg's suggestions
|
2015-10-19 18:04:54 +02:00 |
|
Jan Hendrik Metzen
|
804c1d9042
|
TST Tests for GP regression and classification
|
2015-10-19 18:04:54 +02:00 |