Explainable Hybrid Models Combining Quantum Computing and Deep Learning for Coffee Leaf Disease Detection and Classification

• Autores: Nameer Mezher Bath
• Directores de la Tesis: Enrique Domínguez Merino (dir. tes.)
• Lectura: En la Universidad de Málaga ( España ) en 2026
• Idioma: inglés
• Enlaces
  • Tesis en acceso abierto en: RIUMA
• Resumen

  • This doctoral thesis proposes explainable hybrid models that combine deep learning and quantum computing techniques for accurate detection and classification of coffee leaf diseases. The research focuses on improving classification performance, generalisation ability, and interpretability using convolutional neural networks, transfer learning models, and hybrid quantum–classical approaches.

    Several experiments were conducted using different datasets of coffee leaf images, including balanced and augmented data, to evaluate the effectiveness of the proposed models. Multiple architectures such as CNN, VGG, ResNet, and hybrid quantum neural networks were analysed and compared using various evaluation metrics.

    Explainable artificial intelligence techniques including Grad-CAM, LIME, and SHAP were applied to improve model transparency and to better understand the decision-making process. The results demonstrate that the proposed hybrid framework achieves high accuracy, robustness, and better generalisation compared to traditional machine learning and deep learning approaches, making it suitable for real-world agricultural applications