Inspiration
Diabetes is widespread issue that affects many people in their day to day life. We wanted to create something that would urge people to focus on their health and be conscious of their lifestyle habits.
What it does
Our model takes in inputs about a person, such as age, height and weight, history of some diseases, and yields a percentage of how likely they are to be diabetic.
How we built it
We found a dataset that contains the health information we needed, such as glucose levels, BMI, history of diseases, and more. Then, we found a second dataset that contains the information to estimate a person's A1c levels because that isn't information that is known by everyone. After getting both datasets, we figured out which features were heavily correlated with calculating diabetes and A1c. We then trained our model using a Random Forest Regression Model, and measured the accuracy of the model. After that, we combined both of our models. We also created a website using Streamlit.
Challenges we ran into
Balancing the datasets had many issues because it was originally very uneven. In addition, combining both of the models was also very tricky.
Accomplishments that we're proud of
We are proud of creating a project that will contribute to people's health and be involved with early decisions efforts to lead a better lifestyle.
What we learned
We learned how to train models and test predictive models.
What's next for Diabetes Predictor
We hope to develop more features that will make our product even more advanced and expand our audience to help even more people.
Log in or sign up for Devpost to join the conversation.