Diabetes Prediction Using Machine Learning: A Game-Changer for Healthcare

Predicting diabetes using machine learning involves creating a model that can analyze individuals' various medical and lifestyle features and predict whether they are likely to develop diabetes in the future. This paper introduces the Dual gradient booting (DGB) approach for diabetes prediction. We collected data from fifteen thousand women between the ages of twenty and eighty who were chosen as samples from the outpatient care checkup data of a Taipei City medical (TCM)institution. These ladies were inpatients between 2018 and 2020, 2021 and 2022. Then, the Z-score normalization method is used for normalizing data, which involves scaling or mapping abnormal data to standard data for preprocessing. We have included independent component analysis (ICA) for feature extraction in the process of separating characteristics within a set of measures. The Python platform has been used to develop the suggested strategy. The proposed approach is assessed in terms of precision, accuracy, recall, and F1 score. When compared to existing methods, the suggested methodology (DGB) performed much better and provided Systematic outcomes. Psychiatrists and medical professionals can make quick and precise decisions on diabetes treatment with the aid of our suggested model's ability to forecast diabetes with accuracy in a short amount of time.