Fast and correct variational inference for probabilistic programming: Differentiability, reparameterisation and smoothing
<p>Probabilistic programming is an innovative programming paradigm for posing and automatically solving Bayesian inference problems. In this thesis, we study the foundations of fast yet correct inference for probabilistic programming.</p> <p>Many of the most successful inference t...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2023
|
| Subjects: |