Elsemüller, Lasse
Preview |
PDF, English
- main document
Download (17MB) | Terms of use |
Citation of documents: Please do not cite the URL that is displayed in your browser location input, instead use the DOI, URN or the persistent URL below, as we can guarantee their long-time accessibility.
Abstract
This dissertation scales probabilistic cognitive modeling by advancing amortized Bayesian inference, an emerging combination of simulation-based inference and deep learning. Situated within the broader AI for science paradigm, it develops, evaluates, and applies methodological contributions that enable probabilistic inference for complex models, regardless of likelihood tractability, as well as large data sets. It encompasses three manuscripts that (i) enable amortized Bayesian comparison of hierarchical models, (ii) propose sensitivity-aware amortized Bayesian inference for large-scale sensitivity analyses, and (iii) evaluate unsupervised domain adaptation for addressing sim-to-real gaps caused by model misspecification.
| Document type: | Dissertation |
|---|---|
| Supervisor: | Voß, Prof. Dr. Andreas |
| Place of Publication: | Heidelberg |
| Date of thesis defense: | 22 October 2025 |
| Date Deposited: | 09 Dec 2025 09:40 |
| Date: | 2025 |
| Faculties / Institutes: | The Faculty of Behavioural and Cultural Studies > Institute of Psychology |
| DDC-classification: | 004 Data processing Computer science 150 Psychology 310 General statistics |
