Early Detection of Type 2 diabetes at the pre-diabetic phase using non-invasive parameters
Designed a model that could detect type 2 diabetes at the pre-diabetic phase using the KNN algorithm. Classified HbA1c data into categories of non-diabetic, prediabetic, or diabetic and matched with respective parameters such as BMI and waist circumference, using over a thousand unidentifiable cohorts secondary data. Quality of the model was tested doing a precision and accuracy test.