Advancements in Machine Learning Algorithms for Predictive Analytics in Healthcare Information Systems Management

Abstract

The current studies have critical limitations such as lack of real-world deployment, biases in Electronic Health Records (EHR)-based models, and computational ineffectiveness. This paper proposes an advanced ML framework incorporating transformer-based deep learning architectures, fairness-aware training, privacy-preserving federated learning in order to focus on those challenges. In contrast to existing models which target specific disease classes, the proposed system generalises across chronic and acute conditions, while ensuring scalability in low-resource settings. In addition, the study enhances prediction reliability with the use of real-time knowledge graphs, AI-powered decision support systems, and bias-mitigation strategies. This work uses real-world hospital data to validate the model, creating a practical roadmap for the adoption of AI in healthcare effectively connecting the dots between theoretical progress and real-world clinical practice. The results enhance early detection of diseases, tailored treatment plans and the reduction of health inequalities establishing predictive analytics driven by AI as one of the tools that will change the face of modern medicine.