.. image:: pystat.png
    :height: 20
    :alt: Statistique
    :target: http://www.xavierdupre.fr/app/ensae_teaching_cs/helpsphinx/td_2a_notions.html#pour-un-profil-plutot-data-scientist

Apprendre en environnement contraint
++++++++++++++++++++++++++++++++++++

*(à venir)*

*Lectures*

* `ProtoNN: Compressed and Accurate kNN for Resource-scarce Devices <http://manikvarma.org/pubs/gupta17.pdf>`_
* `Near-optimal sample compression for nearest neighbors <https://papers.nips.cc/paper/5528-near-optimal-sample-compression-for-nearest-neighbors.pdf>`_
* `Joint learning and pruning of decision forests <https://www.kuleuven-kulak.be/benelearn/papers/Benelearn_2016_paper_53.pdf>`_
* `Optimally Pruning Decision Tree Ensembles With Feature Cost <https://arxiv.org/abs/1601.00955>`_
* `Cost-complexity pruning of random forests <https://arxiv.org/pdf/1703.05430.pdf>`_
* `A Compression Approach to Support Vector Model Selection <http://www.jmlr.org/papers/volume5/luxburg04a/luxburg04a.pdf>`_
* `Resource-efficient Machine Learning in 2 KB RAM for the Internet of Things <http://proceedings.mlr.press/v70/kumar17a/kumar17a.pdf>`_

*Modules*

* `EdgeML <https://github.com/Microsoft/EdgeML>`_ :
  entraîner un modèle lorsqu'on a peu de mémoire.