Predictive Models for Non-communicable Diseases: Exploring the Roles of Machine Learning, Deep Learning, Generative AI, and Optimization

Non-Communicable Diseases (also known as Chronic diseases) are long-lasting health conditions that require regular medical care. NCDs such as heart disease, stroke, diabetes, cancer, and arthritis are major public health threats but they are not inevitable. Lifestyle changes and medical care can prevent or manage many of these diseases. Early detection and intervention are essential to prevent chronic diseases from causing death. To effectively combat chronic illness mortality, early prediction becomes critical, enabling proactive disease prevention strategies. The crucial method for this is to integrate Machine Learning (ML) and Deep Learning (DL) into a patient-centric model, thereby satisfying a critical demand within the healthcare system. The primary objective of this survey is to compile a complete set of works published between 2018 and 2023 that deal with the causes and consequences, chronic disease prediction systems such as ML, DL and Generative AI (GAI) models along with their performance parameters, optimization techniques, chronic disease datasets, and data preprocessing, to shed light on the wide range of research conducted. This survey highlighted the significance of features, the selection of independent variables and the amalgamation of multiple algorithms in enhancing both accuracy and NCDs prediction system performance. Consequently, it becomes feasible to diagnose individuals based on their symptoms. This survey concludes with an attempt to highlight potential future approaches that researchers might examine to achieve better prediction of NCDs and systems exploitable by medical professionals and researchers.