A Comparative Analysis of Diabetes Prediction Using Machine Learning and Deep Learning Algorithms in Healthcare

• Published in: 2024 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC) pp. 1 - 6
• Main Authors: Tripathy, Nrusingha, Moharana, Bhimasen, Balabantaray, Sidhanta Kumar, Nayak, Subrat Kumar, Pati, Abhilash, Panigrahi, Amrutanshu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 27-01-2024
• Subjects: artificial neural network; deep learning; Diabetes; diabetes dataset; Machine learning; Machine learning algorithms; Medical treatment; precision; prediction; Prediction algorithms; Predictive models; Programming; recall
• DOI: 10.1109/ASSIC60049.2024.10508008

Hyperglycemia that persists is a concern with diabetes mellitus. As a result, there may be several complications. According to current morbidity rates, 642 million people will have diabetes worldwide by 2040, or one in every ten persons. There is no question that this ominous figure requires further attention. Care must be taken while handling the vast and highly private information the healthcare sector generates. Diabetes mellitus is one of the many deadly illnesses becoming more widespread worldwide. Medical practitioners desire a trustworthy method for predicting diabetes. Various machine learning (ML) approaches are applied to forecast enormous amounts of data. Due to its fast advancement, machine learning is used in many medical science areas. Although it can be difficult to use analytics to forecast results in healthcare in the long run, it may help practitioners make timely decisions about a patient's treatment and well-being based on massive amounts of data. In this work, the diabetes dataset for Pima Indians is considered. The National Institute of Diabetes and Digestive and Kidney Diseases originally stored this dataset. This study compares seven distinct approaches to machine learning using deep neural network models to predict diabetes. The deep neural network model performs more correctly when compared to conventional machine learning methods. To close the expertise gap between datasets and people, this study will assess the dataset from the viewpoint of a dietician and use machine learning utilizing deep neural network techniques. The results show that, in terms of accuracy score, the hybrid Long Short-Term Memory (LSTM) model performed better than the other models.