-
-
Symptoms autocomplete search
-
Symptom tags
-
Diagnosis
Inspiration
As winter approaches, so does the season of flu and sickness, which prevents people from enjoying the holiday season. Illnesses during this season make many people feel tired and lethargic instead of the joy that they should feel in the holidays! With staff shortages in hospitals and new types of illnesses, people don't know what illness they have, and as a result, they aren't able to treat it. To solve this, I created a machine learning website that diagnoses your sickness based on symptoms.
What it does
The website takes in up to 132 symptoms from the user, and uses machine learning to diagnose the illness from 42 possible options. It also tells the user the actions that they should take to recover from their illness.
How I built it
I used a Flask backend and React JS frontend to make the website, and I trained the Random Forest Classifier model on Google Colab in Python! In the frontend, I got the symptoms from the user and sent it to the backend. Then, the machine learning model that I trained made a prediction on the illness using the symptoms, and sent the diagnosis to the frontend to be displayed.
Challenges I ran into
Finding a training and testing dataset for the machine learning model was challenging. I spent many hours looking for one and was debating if I should create my own dataset. Luckily, I was able to find one on Kaggle! Another challenge that I ran into was transforming the dataset so that it could be used by the model!
Accomplishments that I'm proud of
I'm proud that I was able to complete the project in the given time period. Since this was my first hackathon, I wasn't sure if I would be able to complete it on time! I'm also proud that I built something that can make a positive impact and help people! Finally, I'm proud that I learned many new things, which I can use in future hackathons!
What I learned
Through this project, I explored machine learning, which I had not done before. Specifically, I learned how to train and test the Random Forest Classifier model and implement it in the backend. I also learned how to use the Pandas library in Python to transform the datasets. Finally, I learned how to use a Flask backend with a React JS frontend!
What's next for dr. dAIgnoser
In the future, I wish to add more symptoms that the user can input and more illnesses that can be diagnosed/predicted! I also want to create a similar application to help diagnose illnesses in animals!
Log in or sign up for Devpost to join the conversation.