sklearn/doc/roadmap.rst

273 lines
12 KiB
ReStructuredText

.. |ss| raw:: html
<strike>
.. |se| raw:: html
</strike>
.. _roadmap:
Roadmap
=======
Purpose of this document
------------------------
This document list general directions that core contributors are interested
to see developed in scikit-learn. The fact that an item is listed here is in
no way a promise that it will happen, as resources are limited. Rather, it
is an indication that help is welcomed on this topic.
Statement of purpose: Scikit-learn in 2018
------------------------------------------
Eleven years after the inception of Scikit-learn, much has changed in the
world of machine learning. Key changes include:
* Computational tools: The exploitation of GPUs, distributed programming
frameworks like Scala/Spark, etc.
* High-level Python libraries for experimentation, processing and data
management: Jupyter notebook, Cython, Pandas, Dask, Numba...
* Changes in the focus of machine learning research: artificial intelligence
applications (where input structure is key) with deep learning,
representation learning, reinforcement learning, domain transfer, etc.
A more subtle change over the last decade is that, due to changing interests
in ML, PhD students in machine learning are more likely to contribute to
PyTorch, Dask, etc. than to Scikit-learn, so our contributor pool is very
different to a decade ago.
Scikit-learn remains very popular in practice for trying out canonical
machine learning techniques, particularly for applications in experimental
science and in data science. A lot of what we provide is now very mature.
But it can be costly to maintain, and we cannot therefore include arbitrary
new implementations. Yet Scikit-learn is also essential in defining an API
framework for the development of interoperable machine learning components
external to the core library.
**Thus our main goals in this era are to**:
* continue maintaining a high-quality, well-documented collection of canonical
tools for data processing and machine learning within the current scope
(i.e. rectangular data largely invariant to column and row order;
predicting targets with simple structure)
* improve the ease for users to develop and publish external components
* improve interoperability with modern data science tools (e.g. Pandas, Dask)
and infrastructures (e.g. distributed processing)
Many of the more fine-grained goals can be found under the `API tag
<https://github.com/scikit-learn/scikit-learn/issues?q=is%3Aissue+is%3Aopen+sort%3Aupdated-desc+label%3AAPI>`_
on the issue tracker.
Architectural / general goals
-----------------------------
The list is numbered not as an indication of the order of priority, but to
make referring to specific points easier. Please add new entries only at the
bottom. Note that the crossed out entries are already done, and we try to keep
the document up to date as we work on these issues.
#. Improved handling of Pandas DataFrames
* document current handling
* column reordering issue :issue:`7242`
* avoiding unnecessary conversion to ndarray |ss| :issue:`12147` |se|
* returning DataFrames from transformers :issue:`5523`
* getting DataFrames from dataset loaders |ss| :issue:`10733` |se|,
|ss| :issue:`13902` |se|
* Sparse currently not considered |ss| :issue:`12800` |se|
#. Improved handling of categorical features
* Tree-based models should be able to handle both continuous and categorical
features :issue:`12866` and |ss| :issue:`15550` |se|.
* |ss| In dataset loaders :issue:`13902` |se|
* As generic transformers to be used with ColumnTransforms (e.g. ordinal
encoding supervised by correlation with target variable) :issue:`5853`,
:issue:`11805`
* Handling mixtures of categorical and continuous variables
#. Improved handling of missing data
* Making sure meta-estimators are lenient towards missing data,
|ss| :issue:`15319` |se|
* Non-trivial imputers |ss| :issue:`11977`, :issue:`12852` |se|
* Learners directly handling missing data |ss| :issue:`13911` |se|
* An amputation sample generator to make parts of a dataset go missing
:issue:`6284`
#. More didactic documentation
* More and more options have been added to scikit-learn. As a result, the
documentation is crowded which makes it hard for beginners to get the big
picture. Some work could be done in prioritizing the information.
#. Passing around information that is not (X, y): Sample properties
* We need to be able to pass sample weights to scorers in cross validation.
* We should have standard/generalised ways of passing sample-wise properties
around in meta-estimators. :issue:`4497` :issue:`7646`
#. Passing around information that is not (X, y): Feature properties
* Feature names or descriptions should ideally be available to fit for, e.g.
. :issue:`6425` :issue:`6424`
* Per-feature handling (e.g. "is this a nominal / ordinal / English language
text?") should also not need to be provided to estimator constructors,
ideally, but should be available as metadata alongside X. :issue:`8480`
#. Passing around information that is not (X, y): Target information
* We have problems getting the full set of classes to all components when
the data is split/sampled. :issue:`6231` :issue:`8100`
* We have no way to handle a mixture of categorical and continuous targets.
#. Make it easier for external users to write Scikit-learn-compatible
components
* More flexible estimator checks that do not select by estimator name
|ss| :issue:`6599` |se| :issue:`6715`
* Example of how to develop an estimator or a meta-estimator,
|ss| :issue:`14582` |se|
* More self-sufficient running of scikit-learn-contrib or a similar resource
#. Support resampling and sample reduction
* Allow subsampling of majority classes (in a pipeline?) :issue:`3855`
* Implement random forests with resampling :issue:`13227`
#. Better interfaces for interactive development
* |ss| __repr__ and HTML visualisations of estimators
:issue:`6323` and :pr:`14180` |se|.
* Include plotting tools, not just as examples. :issue:`9173`
#. Improved tools for model diagnostics and basic inference
* |ss| alternative feature importances implementations, :issue:`13146` |se|
* better ways to handle validation sets when fitting
* better ways to find thresholds / create decision rules :issue:`8614`
#. Better tools for selecting hyperparameters with transductive estimators
* Grid search and cross validation are not applicable to most clustering
tasks. Stability-based selection is more relevant.
#. Better support for manual and automatic pipeline building
* Easier way to construct complex pipelines and valid search spaces
:issue:`7608` :issue:`5082` :issue:`8243`
* provide search ranges for common estimators??
* cf. `searchgrid <https://searchgrid.readthedocs.io/en/latest/>`_
#. Improved tracking of fitting
* Verbose is not very friendly and should use a standard logging library
:issue:`6929`, :issue:`78`
* Callbacks or a similar system would facilitate logging and early stopping
#. Distributed parallelism
* Accept data which complies with ``__array_function__``
#. A way forward for more out of core
* Dask enables easy out-of-core computation. While the Dask model probably
cannot be adaptable to all machine-learning algorithms, most machine
learning is on smaller data than ETL, hence we can maybe adapt to very
large scale while supporting only a fraction of the patterns.
#. Support for working with pre-trained models
* Estimator "freezing". In particular, right now it's impossible to clone a
`CalibratedClassifierCV` with prefit. :issue:`8370`. :issue:`6451`
#. Backwards-compatible de/serialization of some estimators
* Currently serialization (with pickle) breaks across versions. While we may
not be able to get around other limitations of pickle re security etc, it
would be great to offer cross-version safety from version 1.0. Note: Gael
and Olivier think that this can cause heavy maintenance burden and we
should manage the trade-offs. A possible alternative is presented in the
following point.
#. Documentation and tooling for model lifecycle management
* Document good practices for model deployments and lifecycle: before
deploying a model: snapshot the code versions (numpy, scipy, scikit-learn,
custom code repo), the training script and an alias on how to retrieve
historical training data + snapshot a copy of a small validation set +
snapshot of the predictions (predicted probabilities for classifiers)
on that validation set.
* Document and tools to make it easy to manage upgrade of scikit-learn
versions:
* Try to load the old pickle, if it works, use the validation set
prediction snapshot to detect that the serialized model still behave
the same;
* If joblib.load / pickle.load not work, use the versioned control
training script + historical training set to retrain the model and use
the validation set prediction snapshot to assert that it is possible to
recover the previous predictive performance: if this is not the case
there is probably a bug in scikit-learn that needs to be reported.
#. Everything in Scikit-learn should probably conform to our API contract.
We are still in the process of making decisions on some of these related
issues.
* `Pipeline <pipeline.Pipeline>` and `FeatureUnion` modify their input
parameters in fit. Fixing this requires making sure we have a good
grasp of their use cases to make sure all current functionality is
maintained. :issue:`8157` :issue:`7382`
#. (Optional) Improve scikit-learn common tests suite to make sure that (at
least for frequently used) models have stable predictions across-versions
(to be discussed);
* Extend documentation to mention how to deploy models in Python-free
environments for instance `ONNX <https://github.com/onnx/sklearn-onnx>`_.
and use the above best practices to assess predictive consistency between
scikit-learn and ONNX prediction functions on validation set.
* Document good practices to detect temporal distribution drift for deployed
model and good practices for re-training on fresh data without causing
catastrophic predictive performance regressions.
Subpackage-specific goals
-------------------------
:mod:`sklearn.ensemble`
* |ss| a stacking implementation, :issue:`11047` |se|
:mod:`sklearn.cluster`
* kmeans variants for non-Euclidean distances, if we can show these have
benefits beyond hierarchical clustering.
:mod:`sklearn.model_selection`
* |ss| multi-metric scoring is slow :issue:`9326` |se|
* perhaps we want to be able to get back more than multiple metrics
* the handling of random states in CV splitters is a poor design and
contradicts the validation of similar parameters in estimators,
`SLEP011 <https://github.com/scikit-learn/enhancement_proposals/pull/24>`_
* exploit warm-starting and path algorithms so the benefits of `EstimatorCV`
objects can be accessed via `GridSearchCV` and used in Pipelines.
:issue:`1626`
* Cross-validation should be able to be replaced by OOB estimates whenever a
cross-validation iterator is used.
* Redundant computations in pipelines should be avoided (related to point
above) cf `dask-ml
<https://ml.dask.org/hyper-parameter-search.html#avoid-repeated-work>`_
:mod:`sklearn.neighbors`
* |ss| Ability to substitute a custom/approximate/precomputed nearest neighbors
implementation for ours in all/most contexts that nearest neighbors are used
for learning. :issue:`10463` |se|
:mod:`sklearn.pipeline`
* Performance issues with `Pipeline.memory`
* see "Everything in Scikit-learn should conform to our API contract" above