Significance of Machine Learning Algorithms to Improve Predictive Analytics in Chronic Disease Management through Pharmacogenomics

Abstract

The treatment of chronic diseases is an important concern in global health. Machine learning (ML)-based disease prediction models are becoming more important for making informed medical decisions in light of the paradigm shift towards preventative care. Integrating genetic, pharmacogenomic, personal health, and psychosocial data can greatly assist healthcare practitioners in making treatment-related decisions for patients with chronic diseases. This study utilizes a Deep Convolutional Neural Network-assisted Chronic Disease Management (DCNN-CDM) through pharmacogenomics and an improved predictive analytics model to enable informed real-time decision-making at the point of care. Data augmentation in terms of feature space allows the DCNN model to avoid over-fitting while effectively capturing high-level features submerged in chronic disease datasets. Afterwards, this work suggests an attention-empowered DCNN model to enhance sick case diagnosis accuracy, which augments data regarding sample space, thereby alleviating the class imbalance issue. Electronic health information data mining is now using predictive analytics to determine individuals at risk of acquiring chronic disease problems. The suggested model may aid in the early and accurate diagnosis of chronic diseases. The numerical outcomes demonstrate that the recommended DCNN-CDM model increases the accuracy rate of 98.7%, patient monitoring rate of 97.5%, F1-score rate of 96.3% and predictive performance rate of 95.1% compared to other existing methodologies.