Department of Physics and Computer Science

Permanent URI for this collectionhttps://dspace.njala.edu.sl/handle/20.500.144402/139

Browse

Filters

2025 - 20262

Advanced Search

Filter by

Settings

Author: search.filters.author.Fofanah, Abdul Joseph

Search Results

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    DeepEVD: Integrating Epidemiological data into deep learning frameworks based on spatio-temporal feature learning for EVD forecasting
    (Elsevier BV, 2025-08-01) Fofanah, Abdul Joseph
    The paper introduces DeepEVD, an innovative framework that integrates human mobility data to forecast Ebola Virus Disease (EVD) outbreaks. Traditional epidemiological models often struggle to account for the dynamic nature of human movement, which is crucial for understanding EVD transmission. DeepEVD leverages diverse mobility data sources, including phone records, GPS traces, and social media posts, to extract significant spatio-temporal features. It utilises Graph Convolutional Networks (GCN) and Long Short Term Memory (LSTM) networks to establish connections between mobility patterns and EVD cases across both space and time. The framework was tested on real-world datasets from the 2014-2016 West Africa outbreak and the 2015-2016 Sierra Leone outbreak, demonstrating a 5%-10% reduction in forecasting errors compared to baseline methods. Ablation studies reveal the impact of various data sources and feature extraction methods on accuracy. DeepEVD not only delivers state-of-the-art performance, but it also provides actionable insights for EVD prevention and control. Implementation of the proposed DeepEVD can be accessed here https://github.com/afofanah/DeepEVDMob.
  • Thumbnail Image
    Item
    CL3AN-GNN: curriculum-driven for incremental feature disentanglement with multi-phase attention for class-imbalanced node classification
    (Springer Science+Business Media, 2026-10-01) Fofanah, Abdul Joseph
    Imbalanced node classification in graph neural networks (GNNs) presents a significant challenge, where skewed class distributions lead to biased models that underperform on minority classes. To address this, we propose CL3AN-GNN , a curriculum guided three stage attention network that progressively disentangles features from class agnostic to class specific representations through structured learning phases: Engage , Enact , and Embed . The model begins by Engaging with structurally simpler nodes, characterised by low structural marginality and high feature purity, using uniform attention to build robust foundational representations. In the Enact stage, attention is relaxed to dynamically reweight neighbourhoods and enforce orthogonality between structural and semantic features, initiating feature disentanglement. Finally, the Embed stage specialises attention toward minority class nodes and hard examples, refining discriminative features with adaptive loss weighting. We evaluate CL3AN-GNN on eight imbalanced graph benchmarks spanning social, biological, and citation networks. Experiments demonstrate consistent improvements in accuracy, macro F1, and area under the curve over state-of-the-art methods, with particular gains on minority classes. The curriculum framework yields faster convergence, better generalisation to unseen imbalanced graphs, and improved interpretability via gradient stability analysis and attention correlation maps. This work provides a theoretically grounded and empirically validated approach to imbalanced node classification, advancing curriculum learning in graph representation learning.