QREFM Lab excellently won the First Prize at the Singular Stochastic PDE Learning Competition

Two members of QREFM Lab, Le Duy Anh and Le Tat Dat (both students of DSEB 64B class), under the guidance of Dr. Vuong Van Yen, excellently won the first prize at the Singular Stochastic PDE Learning Competition, a competition organized by University College London (UCL) in collaboration with the Academy of Mathematics and Systems Science. This year's competition brings together more than 70 participating teams from undergraduate to doctoral students, from many of the world's leading universities such as Cambridge University, Imperial College London, University of Edinburgh, and many other international research units.

Stochastic partial differential equations (SPDEs) with noise Space–time is one of the fundamentals models to describe complex physical phenomena such as turbulent flows, superconductivity, or more recently, modeling complex financial systems.QREFM team combines the Neural SPDE framework [Salvi et al., 2022] with Deep Latent Regularity Features [Gong et al., 2023] to learn the mild solution of SPDE singularities. In addition, the team also built Log-Spectral Loss in Fourier space to handle the autocorrelation problem that often appears in these solutions.

With the leading results, QREFM Lab (NEU) and the team that won second place - The Avengers from Cambridge University - will be invited by the organizers to present their research results and participate in an in-depth discussion session. This is considered an opportunity to open up many directions for international scientific cooperation in the future.

Detailed information about the contest: https://hackathon2.deepintomlf.ai/competitions/92/