Data analysis was performed on NHANES glycohemoglobin data contained in the nhgh.tsv file. This includes some basic exploratory data analysis to determine distributions of variables and to determine how the data should be cleaned. SMOTE oversampling was used to balance the dataset before 3 ML models, Logistic Regression, Decision Tree, and Random Forest were used to learn the data and predict if a person has diabetes (gh>=6.5) or not.

Create your own environment and pip install required libraries based on requirements.txt before running the notebook.