Inspiration
After my grandfather got cancer, I figured out he learned it while he was in the 0th stage. The doctors said that it was extremely lucky that he figured out extremely lucky due to a motorcycle accident. After figuring that out, I realized that many other diseases, not only cancer, have been affecting people without them notice. Diabetes, one of the health issues that I have been mainly interested in due to my sweet tooth, I decided I wanted to learn how exactly someone really gets diabetes to see if eating sweetness would be a major risk in order to accomplish my first goal of allowing to help people find if they have diseases.
What it does
By simply using StreamLit to open a new tab, I can identify values for every trait and understand whether with all of these traits together if the person has a diabetes risk.
How I built it
By using the RandomForestCalculator, I was able to create a model by training from a Kaggle file. After that, for the front-end, I used streamlit to easily open a website for it.
Challenges I ran into
One main challenge I ran into was finding a way to maybe a front-end for this. It was a bit challenging considering the fact that I hadn't had much experience with HTML but after finding StreamLit, I was able to use it. In addition, another issue was the constant struggle of syntax where I wasn't sure on whether to put single brackets or double brackets.
Accomplishments that I'm proud of
I'm proud that I was able to connect the front end to the back end and properly show the prediction set worked. In addition, I'm proud that I was able to learn this in a short amount of time as I've struggled with understanding computer science though I'm interested in it.
What I learned
I learned how Machine Learning works by researching and finding ways to train a model. In addition, I was able to learn how to make a simple front-end beginner style that can still bring points.
What's next for Diabetes Prediction
What's next is to create a true front-end though with HTML and all the proper materials.
Log in or sign up for Devpost to join the conversation.