Feature Learning for Bayesian Inference
Description
The goal of this project is to use interpretable Machine Learning (ML) to find low-dimensional features in high-dimensional noisy data generated by (i) stochastic models or (ii) real systems. In both cases, the problem is to disentangle the effect of high-dimensional disturbances, such as noise or unobserved inputs, from the effects of relevant characteristics (model parameters in the first case, system properties in the latter).
Key Data
Projectlead
Prof. Dr. Antonietta Mira
Co-Projectlead
Prof. Dr. Fernando Perez-Cruz
Project team
Dr. Carlo Albert, Prof. Alessandro Laio, Prof. Jukka-Pekka Onnela, Dr. Simone Ulzega
Project status
ongoing, started 09/2022
Funding partner
SNSF
Publications
-
Shedding light on the sun through the calibration of solar dynamo models on millennial records of solar activity
2023 Ulzega, Simone; Albert, Carlo; Beer, Jürg
-
Boosting Bayesian parameter inference of SDE models by Hamiltonian scale separation : a real-world case study in urban hydrology
2023 Ulzega, Simone; Albert, Carlo
-
A comparison of numerical approaches for statistical inference with stochastic models
2023 Bacci, Marco; Sukys, Jonas; Reichert, Peter; Ulzega, Simone; Albert, Carlo
