A collection of PhD theses defended by former doctoral students of the Bio-inspired Computing and Machine Learning (BCML) Lab at Heriot-Watt University and the University of Aberdeen. Theses are indexed in the British Library EThOS (E-Theses Online Service) repository and institutional repositories.
"Automatically Explaining Automated Decision Making for a GDPR User Persona"
Investigates automated natural language generation and explanation techniques for AI-driven decision-making systems tailored to specific GDPR user personas.
"Explainable reasoning for remote autonomous agents"
Focuses on developing explainable AI models, natural language explanations, and uncertainty estimation methods for autonomous navigation and robotic systems.
"Genome engineering using whole-cell modelling and machine learning"
Combines mechanistic biological simulation and machine learning surrogate models to accelerate whole-cell model construction, genome reduction, and E. coli design.
"Schema aware knowledge graph completions"
Investigates schema-aware Knowledge Graph Completion (KGC) methods, consistency checking algorithms, data fusion, and semantic reasoning to balance completeness and consistency in knowledge graphs.
"CLQML-Morven: Towards a Closed-Loop Automated Scientific Discovery System for Dynamic Model Identification in Silico"
Develops a closed-loop qualitative machine learning framework (CLQML-Morven) for automated hypothesis generation and scientific discovery in dynamic biological systems.
"Novel approaches in macro-level neural ensemble architecture search"
Proposes bio-inspired artificial immune network algorithms (ImmuNAS / ImmuNeCS) for searching optimal convolutional neural network architectures and neural ensemble models.
"Operational multistage machine learning fairness"
Investigates operational machine learning fairness, bias quantification, and multi-stage mitigation strategies in algorithmic decision-making systems.
"Prospecting and Informed Dispersal: Understanding and Predicting Their Joint Eco-Evolutionary Dynamics"
Applies individual-based eco-evolutionary modelling and AI tools to study breeding habitat selection and dispersal dynamics in ecological populations.
"Fast hyperparameter optimisation of graph neural network for molecular property prediction"
Focuses on graph neural network architecture hyperparameter optimization, genetic algorithms with tree-structured mutation, and molecular property prediction.
"Clustering analysis on heterogeneous information network based on relation structures"
Develops advanced clustering algorithms and affinity propagation methods tailored for complex heterogeneous information networks (HINs) based on relation structures.
"Investigating rule induction methods in machine learning for improving medical dementia prediction"
Explores interpretable rule induction techniques and bio-inspired classification models for early detection and prediction of dementia from clinical data.
"Inferring Structure and Parameters of Dynamic Systems Using Swarm Intelligence"
Proposes novel particle swarm optimization and evolutionary algorithms for system identification and differential equation parameter estimation.
"Multiple Data Imputation with Clustering Techniques"
Investigates cluster-based data imputation algorithms for handling complex missing data distributions in bio-inspired data analytics.
"Modelling Physarum polycephalum competitive behaviour for solving real world problems"
Models the intelligent foraging and competitive behavior of slime mold (Physarum polycephalum) to design bio-inspired multi-objective optimization algorithms for complex problems.
"Starling swarms optimisation for static and dynamic environments"
Develops starling flocking behavior models and starling swarm optimization algorithms for static and dynamic optimization environments.
"Developmental Learning of Preconditions for Means-End Actions from 3D Vision"
Investigates developmental robotics, cognitive architectures, and reinforcement learning strategies for autonomous robot interaction and problem solving.
