ecfc93ddf6
ENH Add probability calibration based on isotonic regr. and Platt's sigmoid fit + calibration-curve
...
CalibratedClassifierCV allows to calibrate the predicted probabilities of base classifiers based on a cross-validation scheme and either Platt's sigmoid fit or isotonic regression. This can be used to compensate for an under-confident or over-confident classifier. It allows also to turn the decision scores of a non-probabilistic classifier into valid probabilities.
The function calibration_curve allows to evaluate how well calibrated the probabilties returned by a classifier are. Ideally, the curve should be close to diagonal.
2015-02-20 19:37:10 +01:00
f1b42845f4
FIX Add kernel_ridge to list of all submodules
2015-01-18 11:01:43 +09:00
6d7d25fc8c
MAINT dev version to follow PEP440
2014-12-25 23:11:15 +11:00
b4e41cb614
FIX update sklearn.__all__ to include all end-user submodules
2014-09-02 11:23:19 +02:00
be93a75d29
MAINT remove sklearn.test
2014-07-22 15:58:55 +02:00
259b487fb4
MAINT remove deprecated sklearn.pls module
2014-07-22 15:54:54 +02:00
e8cb36f83d
MAINT master to 0.16-git
...
master has 0.16-only commits since the 0.15-branching tag.
New features have to be documented as new features in the 0.16 section.
Bugfixes to be included in the 0.15.0 release have to be cherry-pick'ed in the
0.15.X branch and document in the 0.15 section.
2014-07-14 21:42:06 +02:00
42c31430c9
MAINT: restore -git version in the master branch after the last release merge
2014-06-09 14:53:40 +02:00
bc8666f60f
Branching 0.15.X to prepare the 0.15.0b1 release
2014-06-06 11:57:04 +02:00
61c3b6a32f
FIX use of str.format for Py2.6
2013-11-28 14:53:33 +11:00
9e1edd11e3
FIX stop forcing deprecation warnings for external packages
2013-11-28 14:32:07 +11:00
af5e297302
MAINT/DOC: bump docs and rev numbers to 0.15-git
2013-08-08 01:13:06 +02:00
6bb4cd7d92
TST: avoid nose running sklearn.test as a test
2013-08-07 22:50:33 +02:00
987e2a002a
MISC: deprecation notice
...
Indicate when the function will be removed.
2013-08-06 08:38:30 +02:00
bec85af385
MAINT: remove sklearn.test()
2013-08-06 08:38:30 +02:00
b732f4d1a7
MAINT deprecated the pls module, moved CCA to cca_
...
TEST don't need to excluse _PLS and _CCA
2013-07-25 11:01:59 +02:00
38e84c5fb9
MAINT Move the pls to the cca to a cross_decomposition module
2013-07-25 11:01:59 +02:00
19beac87ff
REL change version to 0.14-git everywhere, update news, support page.
2013-01-21 22:07:26 +01:00
e12c67306e
cosmit pep8
2012-12-16 23:28:10 +01:00
23ac37a4fb
API : move isotonic regression out of linear_model
2012-11-14 22:01:53 +01:00
88f7df2922
COSMIT website address + copyedit in __init__.py
2012-10-23 17:40:33 +02:00
fb7e3abe50
Python3 compat: print()
2012-10-05 17:56:37 +02:00
d50c0cb06c
COSMIT pep8
2012-10-02 17:31:00 +01:00
f1e4427c5c
MISC starting 0.13 cycle
2012-09-04 19:39:19 +01:00
b5519f08da
Merge branch 'linking_arrayfuncs'
...
Conflicts:
setup.py
sklearn/__init__.py
2012-07-26 23:33:29 +02:00
ecda4de954
COSMIT: explainations of the partial import
2012-07-26 11:39:56 +02:00
d8e30f1ca8
REF: hack to be able to share distutils utilities.
...
We took the ugly trick from NumPy to separate build vs normal use of
sklearn imports.
2012-07-25 11:35:40 +01:00
4915d4dcd9
ENH: sklearn.setup_module to preseed RNGs to reproduce failures
...
Could easily be controlled from the environment via SKLEARN_SEED
Besides adding setup_module() to the sklearn space should be of no
effect for any regular imports of sklearn.
2012-07-23 16:42:18 +02:00
cd7404706c
MISC: clean up top-level namespace
2012-05-09 00:32:27 +02:00
67cad73a56
RELEASE HEAD is now 0.12-git
2012-05-07 18:28:34 +02:00
73fdf6a9c9
RELEASE 0.11
2012-05-07 18:19:35 +02:00
df3dac765d
BUG: fix testing on older numpy
2012-05-06 19:37:19 +02:00
1eae414837
TST: do not raise warnings in sklearn.test()
...
This fixes #834
2012-05-06 19:28:19 +02:00
b36beb4e7b
Merge branch 'label-propagation'
2012-02-02 12:55:56 +01:00
268aed6b42
TL directory change to push label propagation code into semi_supervised
2012-01-22 21:13:56 -08:00
4533aa33da
Start of 0.11 development cycle.
2012-01-10 18:27:24 +01:00
a07218757b
Still some tweaks for the sklearn.test() example
2012-01-09 15:15:05 +01:00
2108f11271
FIX: doctest
2012-01-09 14:58:16 +01:00
552bc9d568
DOC: exaple for sklearn.test()
2012-01-09 14:27:30 +01:00
81bff12a09
ENH: Easier debugging in check_build
2011-11-08 22:26:52 +01:00
8782723e3a
MISC: even better check_build error reporting
2011-10-08 19:03:37 +03:00
ea59b84075
MISC: better check_build error reporting
2011-10-08 11:35:51 +02:00
2aaa1fe350
COSMIT
2011-09-24 11:17:38 +02:00
9748a50690
Start of 0.10 development cycle.
2011-09-20 16:13:30 +02:00
54af74d18c
Merge remote-tracking branch 'jakevdp/neighbors-refactor'
2011-09-13 23:41:23 +02:00
a15554d370
Do not import mixture from top-level sklearn.
2011-09-12 22:23:34 +02:00
0718f7231b
merge upstream/master
2011-09-06 20:14:25 -07:00
1583dbbbb4
move neighbors.py -> neighbors
2011-09-06 10:09:35 -07:00
ca48f82ab8
s/\bcross_val\b/cross_validation/g
2011-09-06 14:44:32 +02:00
ddf4b72109
Move project directory from scikits.learn to sklearn
2011-09-02 12:06:57 +02:00
Jan Hendrik Metzen