:orphan: |rss_image| :ref:`<== ` **page de blog - 2/16** :ref:`==> ` :ref:`Blog ` :ref:`article (8) ` :ref:`articles (3) ` :ref:`cours (6) ` :ref:`module (7) ` :ref:`paper (3) ` .. |rss_image| image:: feed-icon-16x16.png :target: ../_downloads/rss.xml :alt: RSS ---- .. _ap-main-1: page de blog - 2/16 +++++++++++++++++++ .. blogpostagg:: :title: Transformation digitale :date: 2021-09-11 :keywords: teachings :categories: conférence :rawfile: 2021/2021-09-11_digital.rst A découvrir... ... .. blogpostagg:: :title: Lecture sur le machine learning et autres... :date: 2021-09-11 :keywords: machine learning :categories: conférence :rawfile: 2021/2021-09-10_ml.rst `Arthur Charpentier, lectures `_ Une conséquence de l'épidémie... .. blogpostagg:: :title: Introduction à la programmation (sketch) :date: 2021-09-08 :keywords: teachings :categories: article :rawfile: 2021/2021-09-08_introduction.rst **Présentation** * Programmation en classes préparatoires, `Informatique en CPGE `_ * architecture d'un ordinateur * base du langage python * arbre binaire, graphes * algorithme, parcours de graphe, distance d'édition, diviser pour régner, programmation dynamique * Langages haut niveau, bas niveau * C/C++ * Rust * Javascript, Python, matlab... * Python : machine learning * datascience, deep learning = python * automatisation, site web * Python populaire (travails, universités) * interfaçage avec C++ * Pratique * Trouver le plus court chemin d'une station de métro à une autre ? --> plus court chemin dans un graphe * Même question en tenant compte des temps de changement de ligne ? ... .. blogpostagg:: :title: Articles à lire, computing :date: 2021-09-02 :keywords: assurance :categories: article :rawfile: 2021/2021-09-02_reading.rst Accelerating deep learning (hardware) * `Google's AI Processor's (TPU) Heart Throbbing Inspiration `_ * `An in-depth look at Google's first Tensor Processing Unit (TPU) `_ Accelerating deep learning (algorithm) * `triton `_ * `deepspeed `_ .. blogpostagg:: :title: Articles à lire, graphes, distance :date: 2021-08-28 :keywords: assurance :categories: article :rawfile: 2021/2021-08-28_reading.rst Quelques liens autour des graphes, distances entre graphes. ... .. blogpostagg:: :title: Articles à lire - JMLR :date: 2021-07-29 :keywords: assurance :categories: article :rawfile: 2021/2021-07-29_reading.rst Assurance, ML * `Predicting Drought and Subsidence Risks in France `_ * `Collaborative Insurance Sustainability and Network Structure `_ * `Autocalibration and Tweedie-dominance for Insurance Pricing with Machine Learning `_ * `A new GEE method to account for heteroscedasticity, using asymmetric least-square regressions `_ JMLR * `Finite-sample Analysis of Interpolating Linear Classifiers in the Overparameterized Regime `_ * `LassoNet: A Neural Network with Feature Sparsity `_ * `Bandit Convex Optimization in Non-stationary Environments `_ * `A General Framework for Adversarial Label Learning `_ * `Non-parametric Quantile Regression via the K-NN Fused Lasso `_ * `Towards a Unified Analysis of Random Fourier Features `_ * `Online stochastic gradient descent on non-convex losses from high-dimensional inference `_ * `Explaining Explanations: Axiomatic Feature Interactions for Deep Networks `_ * `LocalGAN: Modeling Local Distributions for Adversarial Response Generation `_ * `Prediction against a limited adversary `_ * `Sparse Tensor Additive Regression `_ * `RaSE: Random Subspace Ensemble Classification `_ * `Asynchronous Online Testing of Multiple Hypotheses `_ * `FLAME: A Fast Large-scale Almost Matching Exactly Approach to Causal Inference `_ * `Single and Multiple Change-Point Detection with Differential Privacy `_ * `Simple and Fast Algorithms for Interactive Machine Learning with Random Counter-examples `_ * `On Multi-Armed Bandit Designs for Dose-Finding Clinical Trials `_ * `Consistent estimation of small masses in feature sampling `_ * `On Multi-Armed Bandit Designs for Dose-Finding Clinical Trials `_ python * `river `_: online machine learning * `mvlearn `_ * `pykeen `_ * `POT `_ (Python Optimal Transfer) * `giotto-tda `_ (topological machine learning) * `Pykg2vec `_ .. blogpostagg:: :title: Article à lire, assurance :date: 2021-07-21 :keywords: assurance :categories: article :rawfile: 2021/2021-07-21_reading.rst * `Collaborative Insurance Sustainability and Network Structure `_ * `Autocalibration and Tweedie-dominance for Insurance Pricing with Machine Learning `_ * `Estimating Inequality Measures from Quantile Data `_ * `Données Agrégées et Variables Compositionnelles : Note Méthodologique `_ .. blogpostagg:: :title: Article à lire - reinforcement learning :date: 2021-07-11 :keywords: apprentissage par renforcement :categories: article :rawfile: 2021/2021-07-11_reading.rst `Causal Reinforcement Learning using Observational and Interventional Data `_ .. blogpostagg:: :title: DGML, open.data.gouv pour le machine learning :date: 2021-06-10 :keywords: scikit-learn,DGML :categories: cours :rawfile: 2021/2021-06-10_mlgouv.rst Jusqu'à présent, `UCI `_ était le premier réflexe pour obtenir un jeu de données propres et ne nécessitant aucune inscription pour obtenir le jeu comme sur Kaggle. L'Etat français vient de faire de même avec `DGML `_ ou *Data Gouv pour le Machine Learning*. Ce sera dorénavant mon premier réflexe avant de retourner à mes vieilles habitudes. .. blogpostagg:: :title: scikit-learn mooc :date: 2021-06-02 :keywords: scikit-learn,mooc :categories: cours :rawfile: 2021/2021-06-02_scikit-learn_mooc.rst Le `consortium scikit-learn `_ a travaillé à l'élaboration d'un MOOC : `Machine learning in Python with scikit-learn `_. Le contenu offline de ces cours sont aussi accessibles sur github : `Scikit-learn course `_. ---- |rss_image| :ref:`<== ` **page de blog - 2/16** :ref:`==> ` :ref:`2022-10 (1) ` :ref:`2022-12 (2) ` :ref:`2023-01 (1) ` :ref:`2023-02 (1) ` :ref:`2023-04 (1) `