Dataverse Datathon '21 organised by UWaterloo Data Science Club
Given data on patient features such as age, sex, blood pressure, body mass index, physical health, and many others, we seek to predict the presence of diabetes.
The dataset is a cleaned version of the BFRSS 2015 dataset containing 70692 rows and 22 columns.
pandas, seaborn, scikit-learn, matplotlib, and tensorflow were used for data analysis.
To make predictions, support vector machines, decision tree, random forest, and neural networks were used for modelling.