2020 Data Challenges (1)

2020 Challenges

Challenges start on January^1, 2020. An intermediate closing takes place in June with an evaluation of predictions on new test data. The final closing is in December, with a prize-giving ceremony in January 2021.

This year, the challenges were organized and supervised at ENS by Rudy Morel, Tanguy Marchand, Florenting Guth, Louis Thiry, Gaspard Rochette, and John Zarka. The organization of these data challenges is supported by the CFM Chair at the Ecole Normale Supérieure, and by the Fondation des Sciences Mathématiques de Paris.

In this first session, the following 8 challenges are presented:

• " Where will the next trade of a stock occur ? " presented by Eric Lebigot of Capital Fund Management. Given the order book and recent transactions, the aim of this challenge is to predict on which platform the next trade of a given stock will occur.
  
• " Mortality risk prediction presented by Rémy Dubois from Owkin. The aim of this challenge is to predict a patient's survival time from CT scans in the form of 3D images and clinical data in the form of previously extracted descriptors.
  
• " NLP applied to the analysis of legal decisions " presented by Marc Sanselme from Prédilex. Given a case law, the aim of this challenge is to extract the following information from the text: gender of the victim, date of the accident and date of injury consolidation.
  
• " Estimation of industrial production presented by Matthieu Mazzolini from Kayrros. The aim of this challenge is to estimate the production of industrial group sites based on daily measurements and capacity data.
  
• " Multimodal classification of product data presented by Laurent Ach from Rakuten. The aim of this challenge is to perform a multimodal classification based on text and images of articles into product classes.
  
• " " presented by Karl Bartoli from QRT. The aim of this challenge is to predict the returns of a stock on the US market from the returns of the last twenty days.
  
• " AI applied to meter reading " presented by Nicolas Daviaud from Suez. The aim of this challenge is to create an algorithm that reads the value of a meter from images taken under various angles and conditions.
  
• " Metamodels to improve energy consumption and comfort in large buildings presented by Sylvain Le Corff from Oze-Energies. The aim of this challenge is to predict the energy consumption and temperature of a building based on sensor readings.