Diffuser et promouvoir la culture en mathématiques et en informatique déployée dans les sciences agronomiques à INRAE et rassembler la communauté des maths-info INRAE.

Back to top

Data Challenge

28/02/2021 - 02/07/2021
Type d'événement ou fait marquant: 
Evénement

The challenge treats the timely problem of predicting extreme wildfire activity in space and time by using a set of predictor variables, such as meteorological conditions and land cover. Based on a training dataset of more than 500,000 monthly observations of data of wildfire occurrence numbers and aggregated burnt areas over a grid covering the United States and for the years from 1993 to 2015, the participating teams must provide estimations of exceedance probabilities over a range of severity thresholds for both the number of observed wildfires and their aggregated burnt area. Details on the dataset, and detailed instructions on how to participate, can be found here. Intermediate and final rankings of the participating teams will be published on this page.

Unité organisatrice: