Diabetes Prediction System Using Machine Learning

Diabetes is long-lasting disease to causes a health care crisis , but it is challenging to predict without medical tests According to the International Diabetes Federation, there are currently 537 million people living with diabetes worldwide. By the year 2045, this number is projected to double, reaching 783 million. Diabetes is a disorder triggered by the abnormal increase in blood glucose levels. This excess blood glucose produces the symptoms of urination, inceased thirst, and improved hunger. Diabetes is a significant contributor to various health complications, including blindness, kidney failure, amputations, heart failure, and strokes. When we consume food, our body processes it into sugars, or glucose. Machine learning algorithms can process vast amounts of data in medicine, revealing confidential information. The aim of study to investigate how machine learning algorithms can used to predict disease. Decision Trees, Numpy, Support Vector Machine are used to predict the diabetes using machine learning techniques. All these machine learning methods/algorithms can examine vast amounts of data, enabling medical practitioners to make more cost- effective judgments. By analyzing clinical data, establishing connections among multiple diagnostic variables, building a predictive model, and comparing it to a new dataset, the study found that the Support Vector Machine Classifier predicted diabetes illnesses with 80% accuracy. The study found that employing a larger dataset for train and test can improve the model's accuracy and modify different hyperparameters. The study concludes that artificial intelligence using ML modeling can categorize people with or without diabetes.