detection based on nearest neighbors.

:issue:`5279` by `Nicolas Goix`_ and `Alexandre Gramfort`_.

:class:`decomposition.NMF`, allowing the optimization of all

beta-divergences, including the Frobenius norm, the generalized

Kullback-Leibler divergence and the Itakura-Saito divergence.

particularly useful for targets with an exponential trend.

:issue:`7655` by :user:`Karan Desai <karandesai-96>`.

Average Gradient, in :class:`linear_model.LogisticRegression` and

:class:`linear_model.Ridge`. It allows the use of L1 penalty with

multinomial logistic loss, and behaves marginally better than 'sag'

:class:`kernel_approximation.SkewedChi2Sampler`. Since the Skewed-Chi2

kernel is defined on the open interval :math:`(-skewedness; +\infty)^d`,

the transform function should not check whether ``X < 0`` but whether ``X <

- The ``min_weight_fraction_leaf`` constraint in tree construction is now

more efficient, taking a fast path to declare a node a leaf if its weight

solver. :issue:`8835` by :user:`Joan Massich <massich>`.

- Prevent cast from float32 to float64 in

- Add ``max_train_size`` parameter to :class:`model_selection.TimeSeriesSplit`

:issue:`8282` by :user:`Aman Dalmia <dalmia>`.

Bug fixes

interpolates between operating points, and instead weighs precisions

by the change in recall since the last operating point, as per the

`Wikipedia entry <http://en.wikipedia.org/wiki/Average_precision>`_.

(`#7356 <https://github.com/scikit-learn/scikit-learn/pull/7356>`_). By

`Nick Dingwall`_ and `Gael Varoquaux`_.

that produced a singular covariance matrix would cause the helper method

``_c_step`` to throw an exception.

:issue:`3367` by :user:`Jeremy Steward <ThatGeoGuy>`

- Add ``shuffle`` parameter to :func:`model_selection.train_test_split`.

:issue:`8845` by :user:`themrmax <themrmax>`

- Fix bug where stratified CV splitters did not work with

- Fixed a bug in :class:`linear_model.RandomizedLasso`,

where the parameter ``precompute`` were not used consistently accross

classes, and some values proposed in the docstring could raise errors.

:issue:`5359` by `Tom Dupre la Tour`_.

- Fixed a bug where :func:`model_selection.validation_curve`

reused the same estimator for each parameter value.

API changes summary

:issue:`2879` by :user:`Stephen Hoover <stephen-hoover>`.

- The ``decision_function`` output shape for binary classification in

to scikit-learn conventions. :issue:`9100` by `Andreas Müller`_.

- Gradient boosting base models are no longer estimators. By `Andreas Müller`_.

method ``check_decision_proba_consistency`` has been added in

**sklearn.utils.estimator_checks** to check their consistency.

:issue:`7578` by :user:`Shubham Bhardwaj <shubham0704>`

- In version 0.21, the default behavior of splitters that use the

specifying ``train_size`` alone will cause ``test_size`` to be the

remainder. :issue:`7459` by :user:`Nelson Liu <nelson-liu>`.

:class:`cluster.bicluster.SpectralBiclustering` now accept ``y`` in fit.

:issue:`6126` by :user:ldirer

removed in 0.21 due to poor performance.

:issue:`8996` by `Andreas Müller`_.