GithubHelp home page GithubHelp logo

sdpython / ensae_teaching_cs Goto Github PK

View Code? Open in Web Editor NEW
68.0 68.0 59.0 296.67 MB

Teaching materials in python at the @ENSAE

Home Page: http://www.xavierdupre.fr/app/ensae_teaching_cs/helpsphinx3/index.html

License: MIT License

HTML 0.09% Shell 0.01% Python 3.54% C# 0.05% Batchfile 0.01% R 0.01% Jupyter Notebook 94.28% TeX 1.19% PostScript 0.34% JavaScript 0.06% CSS 0.44%
algorithms data-science distributed-computing ensae machine-learning python3 teaching-materials

ensae_teaching_cs's People

Contributors

alexiseidelman avatar ds-anne avatar iwontbecreative avatar rmenuet avatar sdpython avatar xadupre avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

ensae_teaching_cs's Issues

rédiger un ou deux notebook sur le traitement du langage

Aborder la distance d'édition, n-grams, NLTK, gensim,
word2vec, LDA (Latent Dirichlet Application), traduction statistique,
td-idf, coocurrence, analyse de sentiment, stemming
SMT <https://en.wikipedia.org/wiki/Statistical_machine_translation>,
alignement,
moses <http://www.statmt.org/moses/>

contenu, idées futures en vrac

=================
La page des idées

Le problème quand on commence, c'est qu'il n'y jamais de fin.
Quelques idées non traitées qui pourront peut-être intéresser quelques contributeurs.
D'autres sont recensées sur
GitHub/issues <https://github.com/sdpython/ensae_teaching_cs/issues>_
(ou Waffle <https://waffle.io/sdpython/ensae_teaching_cs>_).

.. contents::
:local:

Algorithmes (1A)

.. todoext::
:title: Ajouter un exercice sur le templating (jinja2, mako)
:issue:
:tag: plus

Le templating fonction comme la fonction ``format`` et permet
en plus de faire des répétitions. C'est très utiliser
dans le domaine du web, notamment par `django <https://www.djangoproject.com/>`_.

.. todoext::
:title: ajouter un notebook sur flexx
:issue: 13
:tag: plus

Voir blog post sur :ref:`Flexx <blog-post-flexx>`.

.. todoext::
:title: Estimer le n pourcentile d'une variable aléatoire
:tag: plus

Intéressant pour une séance de travaux pratiques.
`Estimating the n percentile of a set <http://matthewfl.com/2018/programming/algorithms/estimating-the-n-percentile-of-a-set>`_

.. todoext::
:title: Simuler différentes systèmes de votes
:tag: plus

Intéressant pour une séance de travaux pratiques.
`Theoretical online voting system <http://matthewfl.com/2094/programming/algorithms/theoretical-online-voting-system>`_,
`Survey of Fully Verifiable Voting Cryptoschemes <https://courses.csail.mit.edu/6.857/2016/files/2.pdf>`_.

Data Science (2A)

Programmation

.. todoext::
:title: ajouter un notebook sur joblib
:tag: plus

joblib est utilisé par scikit-learn pour
paralléliser les calculs

.. todoext::
:title: Retravailler la partie visualisation de Python pour un data scientist
:tag: plus

Il manque un notebooks sur les visualisations les plus utilisées en machine learning,
ROC, régression, visualisation d'arbres de décision avec ete3, les cartes.
Insister sur l'interactivité.
Voir `TD 4B : Visualisation <http://www.xavierdupre.fr/app/actuariat_python/helpsphinx/notebooks/seance6_graphes_enonce.html#seance6graphesenoncerst>`_
(`correction <http://www.xavierdupre.fr/app/actuariat_python/helpsphinx/notebooks/seance6_graphes_correction.html#seance6graphescorrectionrst>`_),
ce notebook présente un moyen de faire une carte géographique, des graphes zoomables.

.. todoext::
:title: ajouter ctypes
:tag: plus

utilisation du module ctypes pour les import C++
+ un exemple de `sklearn-compiledtrees <https://github.com/ajtulloch/sklearn-compiledtrees/>`_

Programmation avancée

.. todoext::
:title: aborder d'autres librairies
:tag: plus

py-earth, pytorch, boruta, wendelin.core, zodb,
(requires transaction, zc.lockfile, zodbpickle, ZODB, zdaemon, ZEO, ZODB3, wendelin.core),
ghost.py (scrapping)
h5py, PyTables, lda
See `Related Projects <http://scikit-learn.org/stable/related_projects.html>`_,
`Python extensions to do machine learning <http://www.xavierdupre.fr/blog/2013-09-15_nojs.html>`_

.. todoext::
:title: ajouter un notebook sur numba, llvmlite
:tag: plus

Il n'y pas que CPython pour ooptimiser les calculs.
Aborder les notions de JIT.

.. todoext::
:title: ajouter MILP
:tag: plus

avec des modules tels que `pyomo <http://www.pyomo.org/>`_,
lire `Mixed integer programming for machine learning <http://www.litislab.fr/wp-content/uploads/2015/12/Canu-S.pdf>`_,
`GLPK/Python <https://en.wikibooks.org/wiki/GLPK/Python#Python-GLPK>`_,
`optlang <http://optlang.readthedocs.io/en/latest/>`_

Théorie

.. todoext::
:title: multi-label, coverage_error
:tag: plus

fonction `coverage-error <http://scikit-learn.org/stable/modules/model_evaluation.html#coverage-error>`_,
lire `Mining Multi-label Data <http://lpis.csd.auth.gr/publications/tsoumakas09-dmkdh.pdf>`_

aborder d'autres librairies

py-earth, pytorch, boruta, wendelin.core, zodb,
(requires transaction, zc.lockfile, zodbpickle, ZODB, zdaemon, ZEO, ZODB3, wendelin.core),
ghost.py (scrapping)
h5py, PyTables, lda
See Related Projects <http://scikit-learn.org/stable/related_projects.html>,
Python extensions to do machine learning <http://www.xavierdupre.fr/blog/2013-09-15_nojs.html>

ajouter une séance sur le calcul de la note des aliments

La formule est explicitée sur Open Food Facts <http://fr.openfoodfacts.org/score-nutritionnel-france>. L'objectif est de l'implémenter. Les données sont disponibles sur data.gouv.fr <https://www.data.gouv.fr/fr/datasets/produits-alimentaires-ingredients-nutrition-labels/>.

Retravailler la partie visualisation de Python pour un data scientist, ajouter dash

Il manque un notebooks sur les visualisations les plus utilisées en machine learning,
ROC, régression, visualisation d'arbres de décision avec ete3, les cartes.
Insister sur l'interactivité.
Voir TD 4B : Visualisation <http://www.xavierdupre.fr/app/actuariat_python/helpsphinx/notebooks/seance6_graphes_enonce.html#seance6graphesenoncerst>_
(correction <http://www.xavierdupre.fr/app/actuariat_python/helpsphinx/notebooks/seance6_graphes_correction.html#seance6graphescorrectionrst>_),
ce notebook présente un moyen de faire une carte géographique, des graphes zoomables.

techniques de webscrapping

  • beautifulsoup, ghost.py, scrappy
  • Python's Web Framework Benchmarks <http://klen.github.io/py-frameworks-bench/>_

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.