Paulius Dilkas

Research project title: 
Probabilistic Inference via Weighted Model Counting: Algorithms, Encodings, and Random Instances
Research project: 

The goal of this project is to better understand and improve the performance of inference in probabilistic graphical models such as Bayesian networks and probabilistic relational models such as probabilistic logic programming languages. We aim to improve our understanding of inference with synthetic data generation and parameterised complexity. Performance improvements are sought by focussing on the computational task behind many inference algorithms--weighted model counting (WMC). We consider ways to perform WMC more efficiently by setting it in a more expressive format than that of propositional formulae in conjunctive normal form.

Supervisor: 
Student type: 
Current student