Programme détaillé

• 09:50 - Machine Learning for Computer Security Experts using Python & scikit-learn

  Anaël Bonneton, ANSSI - ENS Paris

  We present SecuML, a Python open source tool that aims to foster the use of Machine Learning in Computer Security. It allows security experts to train models easily and comes up with a user interface to visualize the results and interact with the models.

• 10:30 - Cloud computing made easy in Joblib

  Alexandre Abadie, Inria
  

  Joblib is a Python package initially designed for efficient computing of embarrassingly parallel problems on a local computer or a laptop. This talk gives a short introduction of the features provided by Joblib and the recent developments that make them usable on Cloud computing infrastructures.

• 10:50 - Imbalanced learn

  Guillaume Lemaître, Inria Saclay

  Imbalanced-learn, a Python module to perform under sampling and over sampling with various techniques.

• 11:10 - Introduction to pomegranate

  Venkat Raghav Rajagopalan, Telecom Paristech

  Pomegranate is a python module for probabilistic modelling focusing on both ease of use and speed, beating out competitors in benchmarks. In this talk I will describe how to use pomegranate to simply create sophisticated hidden Markov models, Bayesian Networks, General Mixture Models (and more!).

• 14:00 - Building and deploying a predictive API using scikit-learn, Flask and Docker

  Nawfal Tachfine, Aramisauto.com

  The aim of this workshop is to expose a trained scikit-learn machine learning model as a REST API, built with Flask and Docker, to be queried by any system in JSON.

• 14:40 - How to prepare your text data for NLP applications

  Loryfel Nunez, FindSignal

  Data cleaning is not as sexy as the the actual NLP algorithms. True. BUT, how your prepare your data will determine how well, or poorly your algorithm will perform. This talk will focus on Python’s libraries to extract important text or to remove unwanted text to prepare your data for NLP tasks.

• 15:20 - Machine Learning to moderate ads in real world classified’s business

  Vaibhav Singh, OLX Naspers Services GmbH

  In this talk we share our experiences on how we at OLX Berlin built machine learning models to moderate 100+ million classified ads every month. Audience will get a chance to experience a real world of content moderation and a race to beat online fraudsters and scammers.