Kasper Bågmark

I am a PhD student studying the nonlinear filtering problem using tools from Bayesian statistics and stochastic analysis. I explore ways to overcome the curse of dimensionality, currently focusing on deep learning-based methods to solve the filtering equations via PDEs and SPDEs.

I am particularly interested in collaborative projects related to state-space models, inference for conditional distributions, and numerical schemes for high-dimensional partial and stochastic differential equations.

The annual Dynstoch Workshop will take place in Gothenburg in June 2026 — we’re excited to be hosting it! Check out the webpage for the latest updates.

Proposals and collaboration

We are always looking for good Master’s thesis students for projects in nonlinear filtering and associated areas. Currently, we are investigating deep BSDE methods for nonlinear filtering, here is a proposal for a motivated student!

Here you see the master’s thesis of Filip Rydin who worked on a similar problem who was supervised by me and Adam Andersson.

If you’re interested in collaborating or have a proposal related to probabilistic learning, Bayesian statistics, or differential equations (PDEs, SPDEs, or SDEs), feel free to get in touch!