Med Assist Bot: AI Based Diabetes Prediction Tool for Assisting Novice Medical Practitioners

The decision accuracy of a person is adequate with a small amount of data and experience, but as data increases with different aspects and relationships, the decision accuracy decreases. However, this is not the case with machines. The emergence of artificial intelligence and integration with the medical field, the decision accuracy is increased and speedup. The aim of the study is to show the significance of artificial intelligence in the medical field and to achieve the aim a case study of diabetes prediction system using machine learning is implemented. In the study, a diabetes dataset and five different machine learning algorithms (logistic regression, k-nearest neighbours, decision tree, random forest and gradient boosting) are used. Since, high accuracy of the diabetic prediction system is required; a random over sampling technique is also used to improve the performance of the system. The diabetic prediction system claimed an accuracy of upto 87%.