The data science competition was open to students from IMT Schools from October 2 to December 22, 2017, and featured a challenge on “Well Performance Prediction”: a first for the Institute, organized by Mines Saint-Étienne and Total. Three engineering students from Mines Saint-Étienne and IMT Lille-Douai are on the podium!
The challenge, led by Olivier Roustant (Mines Saint-Étienne, Institut Fayol) and Michel Lutz (Total), was a great success among students with over 200 participants, and more than 2,500 contributions recorded on the site.
The objective was to provide the best production forecasts for an oil well based on a dataset, in order to optimize its operation.
The ranking of candidates was carried out according to a two-step procedure:
- Selection of the 10 best students based on the predictive quality of their model.
- Evaluation of a report explaining the approach, assessed by a jury composed of IMT faculty and Total representatives.
We extend our warmest congratulations to the three award-winning students:
1st – Naoufal ACHARKI, Mines Saint-Étienne, ICM 2nd year
2nd – Martial GARCHERY, engineering student at IMT Lille-Douai
3rd – Julien COLOMB, Mines Saint-Étienne, ICM 2nd year
They each received an internship offer in data science at Total.
A special prize is also awarded to the random forest method, which was adapted in various forms by the three winners and combined with appropriate preprocessing of the dataset.
