SigMA
Implementation and experiments for learning parameters in fractional-Brownian-motion-driven SDEs using path signatures and multi-head attention.
Software & resources
Original repositories accompanying selected publications—connecting mathematical methods, numerical experiments, and the results they support.
Open research
These repositories contain implementations and experiments developed with collaborators across computational finance, stochastic systems, machine learning, and environmental uncertainty.
Each card links to the original public repository maintained by the project owner—not to a fork—and pairs it with the associated paper and research theme.
Research software library
Seven codebases, organized by application area and connected to the research they reproduce.
Data-driven tools for parameter learning in stochastic models and for estimating climate-related extremes.
Implementation and experiments for learning parameters in fractional-Brownian-motion-driven SDEs using path signatures and multi-head attention.
Training, prediction, evaluation, and visualization workflows for neural-network storm-surge models using atmospheric and tide-gauge data.
Reproducible implementations of Fourier and randomized quasi-Monte Carlo methods for pricing and multivariate risk.
Numerical experiments for single- and multi-level Fourier randomized quasi-Monte Carlo methods applied to multivariate shortfall risk.
Implementations of randomized quasi-Monte Carlo, Monte Carlo, and Gauss–Laguerre quadrature in Fourier space for multi-asset options.
Code accompanying the bachelor thesis on numerical uncertainty quantification in option pricing.
Research code for filtering, dimension reduction, rare-event estimation, and efficient multilevel simulation.
MATLAB implementation of standard and filtered Markovian projection methods for dimension reduction in partially observed reaction networks.
MATLAB routines comparing multilevel estimators with and without importance sampling for stochastic biological systems and rare events.
Explore the context
For article links, preprints, and citation details, visit the full publication record. For the questions and methods connecting these projects, explore the research hubs.