Example MSC Theses

• Detection of Overshooting Tops in Deep Convective Clouds using Deep Learning
• Description Logic Embeddings for Neuro-Symbolic Reasoning
• New Benchmarks for Knowledge Graph Completion
• A Formalization And Attack Tree for Gasper: Ethereum's Proposal for Proof-of-Stake Consensus
• Video Annotation with Knowledge Graphs, using Deep Learning and Background Knowledge
• Experimental Comparison of Ansatze for Quantum Natural Language Processing
• Learning Martingales
• Improving the Relational Inductive Bias of Graph Neural Networks
• Assessing the Interpretability of Large Language Models
• Developing computational tools to aid the design of CRISPR/Cas9 gene editing experiments
• Using State-of-the-art Computer Vision Models to Build Novel Applications on Recent XR Platforms
• Neural network architectures to identify risk of Parkinson’s disease
• A unique normal form for prime-dimensional qudit Clifford ZX-calculus
• Reinforcement Learning via Predictive Coding: Revisiting Perspectives from Neuroscience in the Context of Deep Learning
• Motion Planning for Underwater Gliders
• Enhancing VAE-learning on spatial priors using graph convolutional networks
• Deep Learning for Alzheimer’s Disease Genomics
• Machine Learning Approaches for Estimating the Causal Effects of Aerosols on Clouds
• Structural Transfer Learning in Semantic Parsers
• Augmenting semi-mechanistic models with deep learning to model the SARS-CoV-2 pandemic
• Beyond MCMC: A Variational Inference Workflow
• Graphical calculus for Tensor Network Contractions
• Quantum Circuit Optimisation Through Stabiliser Reduction of Pauli Exponentials
• Portable, Low-Cost, and Low-Power mmWave Radar Imaging for Concealed Weapon Detection
• Participatory Budgeting with Conflicts and Partial Information
• Essays on Participatory Budgeting
• Assessing the performance and security characteristics of QUIC in the context of computationally constrained spacecraft
• Quantum Machine Learning using the ZXW-Calculus
• Compiling Deep Learning Workloads for Distributed Computing Clusters
• Category-theoretic syntactic models of programming languages

• Mixed Integer Linear Programming for the Train Timetabling Problem
• Developing robust machine learning models for off-target prediction in CRISPR/Cas9 gene editing using ensemble learning and feature space analysis
• Dependent Types with Borrowing
• Imitate Human in Board Games: Move Prediction in Chess and Xiangqi with Structurally Variable Networks
• Task-adaptive Continual Learning
• On Representation Learning for Deterministic Uncertainty Estimation
• Conformal time-series forecasting
• SiPPI - Predicting Protein-Protein Interactions using Deep Learning
• Deep learning for tabular data in healthcare
• Analysis and Explanation of Binarized Neural Networks using Symbolic Trajectory Evaluation
• Hypergraph MBQC in the ZH-calculus
• Associative Memories and Predictive Coding
• Quantum Computing for Graph Representation Learning
• Exploring the latent space of deep generative protein models: Applications to GPCRs
• Automated Classification of Cardiac Action Potential Phenotypes for Prediction of Drug-Induced Pro-Arrhythmic Risk
• Machine Learning in Particle Physics: Graph Neural Networks for Jet Clustering at the Future Circular Collider (FCC-ee)
• Self-supervised Single-Image 3D Face Relighting
• An Empirical Investigation of Training Speed and Generalisation
• Crafting Label Noise to Increase Adversarial Vulnerability in Deep Learning Models
• Nowcasting unemployment with real time job vacancies data in the United States
• Deep Learning models for the analysis of short tandem repeats in DNA sequences
• Security for Health An Exploration of Cybersecurity, eHealth, and Sustainable Development Goal Three
• Software implementation of Ethical Black Box, a software, which can be used to collect data about a robot’s actions in real-time and in context, to recreate and analyse its past actions
• DeDUCE: A Novel Algorithm Providing Counterfactual Explanations
• Voting and Strategic Candidacies in Primary Systems
• 4-Player Chess: An Environment and Benchmarks for Multi-Agent Reinforcement Learning
• Intent and Theory of Mind: Modelling Belief under Uncertainty
• Moving Object Detection Based On Weakly Supervised Deep Learning Method
• Novel Predictive Coding Networks that Achieve Task-Agnostic Learning
• Inference Attacks on Graph Neural Networks
• Learning to play Open Face Chinese Poker through Deep Reinforcement Learning
• Inferring Dynamic Brain Networks with RNN-VAEs on MEG Data
• Persus: A Latency and Fault Tolerance Library for Microservices Architecture in Golang
• A Box Embedding Model for Temporal Knowledge Graph Completion
• Optimizing large scale omics analyses with data management techniques
• Multi-Type Continuous Disentanglement Variational AutoEncoder Projecting multiple categorical models into a single dimensional space
• Evaluating MAP Inference for Chase-Generated Hinge-Loss Markov Random Fields
• Conversational Negation in Compositional Distributional Semantics
• Variability in Drug Response: A Comparison of Population of Models and Uncertainty Quantification for Predictions of Drug-induced Effects on the Cardiac Action Potential
• On Representation Learning for Deterministic Uncertainty Estimation
• Improving weight-sharing Neural Architecture Search with a marginal likelihood estimator
• Reinforcement Learning for Rule Selection in End to End Dierentiable Proving
• Visualising Simulated Liquid Democracy and Measuring Performance Under Changing Agent Parameters
• Cheap Talk Discovery and Utilization in Multi-Agent Reinforcement Learning