Using Structured Priors when Analysing Image Data

Machine learning is a key technique in many different areas, and machine learning accounts for many of the recent successes in artificial intelligence.

Data is however scarce in most applications, which is why well-posed priors and penalties have been very important in reigning in the often high-dimensional problems considered.

Structured priors and penalties take this a step further, by not just penalising single variables in isolation but penalising deviations from particular structures.Structured priors opens up a toolbox of means to encode general domain-specific knowledge into a machine learning model.

Current means to encode prior information do however either not allow structured relationships, do not guaranteed to draw samples close to the true posterior, have sub-optimal convergence rates, or do not provide uncertainty estimates of the parameters or of the model predictions.This project proposes to resolve all these issues, and provide a generic modelling framework with potential applications in many areas of medicine, science, and technology.

We will develop novel structured priors and sampling algorithms to improve interpretability, variable selection, and uncertainty estimation in machine learning.

The methods will be evaluated in medical imaging applications (reconstructing quantitative magnetic resonance images and predicting schizophrenia, bipolar disorder, and Alzheimer´s disease). However, the proposed methodology is applicable in many diverse areas.