.. blogpost::
    :title: Performance
    :keywords: linear algebra, CPU
    :date: 2021-05-03
    :categories: performance

    Quelques articles sur la performance.

    * `How to optimize GEMM on CPU <https://tvm.apache.org/docs/tutorials/optimize/opt_gemm.html#sphx-glr-tutorials-optimize-opt-gemm-py>`_
    * `SLIDE : In Defense of Smart Algorithms over Hardware Acceleration for Large-Scale Deep Learning Systems <https://arxiv.org/abs/1903.03129>`_
    * `Scalable and Sustainable Deep Learning via Randomized Hashing <https://arxiv.org/abs/1602.08194>`_
    * `Accelerating SLIDE Deep Learning on Modern CPUs: Vectorization, Quantizations, Memory Optimizations, and More <https://arxiv.org/abs/2103.10891>`_

    Other papers:

    * `ZeRO: Memory Optimizations Toward Training Trillion Parameter Models <https://arxiv.org/abs/1910.02054>`_

    Fairness:

    * `How Facebook got addicted to spreading misinformation
      <https://www.technologyreview.com/2021/03/11/1020600/facebook-responsible-ai-misinformation/>`_
    * `Fairness and machine learning <https://fairmlbook.org/>`_

    Neural networks:

    * `Neural Tangent Kernel: Convergence and Generalization in Neural Networks
      <https://arxiv.org/pdf/1806.07572.pdf>`_

    Arbres :

    * `Unbiased Measurement of Feature Importance in Tree-Based Methods <https://arxiv.org/abs/1903.05179>`_

    Collèges de France

    * `Une vision mathématique du "Deep Learning"
      <https://www.college-de-france.fr/site/pierre-louis-lions/seminar-2018-11-09-11h15.htm>`_
    * `Stéphane Mallat, Représentation Parcimonieuse
      <https://www.college-de-france.fr/site/stephane-mallat/course-2021-01-13-09h30.htm>`_
    * `Algorithmes quantiques : quand la physique quantique défie la thèse de Church-Turing
      <https://www.college-de-france.fr/site/frederic-magniez/inaugural-lecture-2021-04-01-18h00.htm>`_

    Autres

    * `Invariant Risk Minimization <https://arxiv.org/abs/1907.02893>`_  (... Bottou ...)
    * `Deep Neural Networks Motivated by Partial Differential Equations <https://arxiv.org/abs/1804.04272>`_
    * `Lagrangian Neural Networks <https://arxiv.org/abs/2003.04630>`_

    ML

    * `A Hierarchical Model for Data-to-Text Generation <https://arxiv.org/abs/1912.10011>`_

    Apprentissage par renforcement

    * `Dynamic Programming in Distributional Reinforcement Learning
      <https://hal.archives-ouvertes.fr/hal-03168889>`_
    * `A Distributional Perspective on Reinforcement Learning
      <https://arxiv.org/abs/1707.06887>`_

    Ethique

    * `Quelle responsabilité pour les algorithmes
      <https://freakonometrics.hypotheses.org/files/2020/06/Risques-122_Bigot_Charpentier.pdf>`_

    Quelques autres directions :

    * `plaidml <https://github.com/plaidml/plaidml>`_,
      `Stripe: Tensor Compilation via the Nested Polyhedral Model
      <https://arxiv.org/abs/1903.06498>`_

    Une énième implémentation de pipeline de machine learning :
    `kedro <https://github.com/quantumblacklabs/kedro>`_.

    Quelques PR intéressantes (scikit-learn) :

    * `Improve near constant feature detection in scalers
      <https://github.com/scikit-learn/scikit-learn/issues/19898>`_
    * `MNT Avoid catastrophic cancellation in mean_variance_axis
      <https://github.com/scikit-learn/scikit-learn/pull/19766>`_