Advances and Challenges in Machine Learning for Diabetes Prediction: A Comprehensive Review

Abstract. Diabetes mellitus is a prevalent and severe metabolic disorder disease that poses significant health risks globally, leading to substantial healthcare burdens. Recent days, advancements in artificial intelligence (AI) have markedly enhanced the accuracy and efficiency of diabetes outcome predicted by machine learning (ML), offering a promising approach for early intervention and treatment. This paper evaluates several advanced ML models, including Random Forest (RF), Support Vector Machine (SVM), and Neural Networks techniques based on neural networks. Each model's strengths and limitations are discussed, highlighting the improvements in predictive performance and diagnostic precision. Despite these advancements, the field faces ongoing challenges related to ethical considerations and data scale, which impact ML application in healthcare from both technical and moral aspects. Future efforts should focus on these challenges by promoting data sharing and integration while safeguarding privacy. Through these endeavors, we aim to advance the field of diabetes prediction and improve patient care.