A doctoral dissertation at the College of Computer Science and Information Technology at University of Basrah examines models and methods for intelligent data analysis in medical diagnosis and pharmacy.
The dissertation, presented by doctoral student Mohammed Monis Ismail Al-Saghir, aims to predict adverse drug reactions (ADIs) using intelligent data mining techniques. The proposed NNPS model combines convolutional neural networks (CNNs) with negative swarm optimization (SSA) to improve prediction accuracy.
The dissertation included data preprocessing, which included feature encoding and balancing. Experimental results demonstrated that NNPS outperformed traditional models in AUROC and AUPRC scores. The study provides a reliable tool for safer polypharmacy management.
