Projects
Research and applied work — from numerical analysis to machine learning in production settings. Each page gives the problem, the approach and the outcome.
Reinforcement learning for Formula 1 race strategy
Internship at the Alfa Romeo F1 Team ORLEN, improving an AI agent that recommends live race-strategy decisions during Grands Prix.
Read more →Uncertainty-aware Bayesian inversion for tumor growth
Calibrating tumor-growth PDE models to patient data with quantified uncertainty, made tractable by a latent-dynamics neural surrogate.
Paper in preparation
Read more →Scalable natural gradient descent for neural PDE solvers
Making natural gradient training of physics-informed neural networks fast, using matrix-free solvers with randomized low-rank preconditioning.
To appear in Computer Methods in Applied Mechanics and Engineering
Read more →Are neural networks a manifold?
Proving that shallow tanh networks, once the obvious redundancies are removed, form a smooth manifold — the geometry behind natural gradient methods.
Paper in preparation
Read more →Low-rank tensor trains for Gross–Pitaevskii ground states
Computing ground states of Bose–Einstein condensates with Riemannian optimization on manifolds of fixed-rank tensor trains.
Read more →Low-rank Riemannian optimization for matrix equations
Solving large multiterm linear matrix equations directly on the manifold of fixed-rank matrices, with new preconditioners for Riemannian CG.
Read more →Neural networks meet isogeometric analysis
Quasi-interpolating neural networks onto spline spaces to combine the flexibility of neural PDE solvers with the accuracy and structure of IGA.
Paper in preparation
Read more →U-Net segmentation of nerve histologies
A convolutional neural network that segments fascicles in vagus nerve histologies, to support more selective nerve stimulation therapies.
Read more →