Zia Ur Rehman
PhD Researcher in Computer Science
Zia Ur Rehman is a PhD researcher at the University of Limerick, working within the Biocomputing and Development Systems (BDS) Group and Lero under the supervision of Prof. Conor Ryan. His research develops privacy-preserving and interpretable machine learning frameworks that integrate ensemble methods using Grammatical Evolution in federated architectures, aligned with responsible AI and computational efficiency in distributed environments.
Alongside his doctoral work, he has published on the energy footprint of machine learning models, contributing to the growing field of Green AI, and maintains open-source tooling for measuring the energy consumption of ML workloads.
COGs-FL: Class-Optimised GANs for Federated Learning under Class Imbalance and Data Heterogeneity
Available at SSRN 7181663.
View publication →Analyzing the energy consumption of random forest and support vector machine models: paving the way for green and sustainable artificial intelligence
Discover Internet of Things, 5, 100.
View publication →Energy-Efficient AI for Medical Diagnostics: Performance and Sustainability Analysis of ResNet and MobileNet
Studies in Health Technology and Informatics, 327, 1225–1229.
View publication →Enhancing Brain Tumor Classification Using CNN: Impact of Data Augmentation on Model Generalization and Performance in Federated Learning
26th International Multi-Topic Conference (INMIC), 1–6. IEEE.
View publication →Improving CNN model training time efficiency using MPI-driven parallelization and ensemble learning
International Conference on AI and the Digital Economy (CADE 2024), 128–133. IET.
View publication →