Data Challenge: a 2025 season extended to 2026 and a new challenge signed by QRT

Data Challenge: a 2025 season extended to 2026 and a new challenge proposed by QRT

The initiative Challenge Data is entering a particular phase in its development.
La Season 2025 is exceptionally extended to the whole of 2026, offering participants the opportunity to continue working on challenges from previous seasons, which remain accessible on the platform.

This transition period marks an adaptation of the calendar before a Return to normal operation as early as 2027, with the launch of a new season including a dozen original challenges.

In this context, Qube Research & Technologies (QRT) proposes a new challenge called “Predicting asset allocation performance”. A historical partner of Challenge Data, QRT is thus continuing its commitment by helping to strengthen the position of Quantitative finance and applied data science among the themes addressed.

Looking back on the 2025 season

As this transition period begins, it is also an opportunity to look back at the 2025 season, closed on December 15.
This edition once again highlighted the diversity of topics covered and the wealth of issues proposed, covering in particular the fields of transport, energy, energy, finance, insurance, health, industry and artificial intelligence.

The challenge partners for the 2025 season were:
Transilien SNCF Voyageurs, Enedis, Crédit Agricole Assurances, Crédit Agricole Assurances, Valeo, SLB, Bertin Technologies, Egis, HrFlow.ai, Raidium, QRT and the Paris Brain Institute (ICM).

The participants, coming from a variety of academic and professional backgrounds, stood out for the scientific quality and operational relevance of their approaches. This year, the valorization of the results resulted in the highlighting of the best contributions directly on the pages of each challenge.

Challenge Data: an open platform for learning, experimenting and innovating

Challenge Data is a platform for challenges in data science open to all: students, researchers, professionals, teachers or people wishing to learn about the discipline.
Each challenge is based on real data, provided by companies, public services or research laboratories, and invites participants to develop rigorous predictive models, evaluated according to transparent scientific protocols.

Traditionally, the seasons are launched in January, with the presentation of new challenges as part of the Stéphane Mallat's course at the Collège de France, and end with an award ceremony associated with this course.

The platform offers:

• Of Participating accounts, to compete individually or as a team;
• Of teacher accounts, making it possible to integrate challenges into educational systems;
• Of challenge provider accounts, for partners who want to propose datasets and monitor the results.

Enrollment is free. In an educational setting, participants may be required to write a scientific report, mainly used by teachers as an assessment medium.

An academic and partnership initiative

Since this year, the Challenge Data platform is managed by the Louis Bachelier Institute (ILB), in close collaboration with its academic partners.
The project is carried out in collaboration with the Collège de France, PSL, ENS Paris, the ILB Data Lab, as well as with the support ofPRAIRIE Institute And of theIDRIS (CNRS), guaranteeing a high level of scientific and technical requirements.

To discover the challenges, participate or propose a project :
Go to the Challenge Data website (registration and guides available online):
www.challengedata.ens.fr

‍