DoctorAI

• 

  Home Page

• 

  Disease Prediction Page

• 

  Choosing Symptoms

• 

  Disease Prediction by ML model

Inspiration

The lack of availability of Doctors (especially in rural areas) and Emergency Situation (especially at night, when it's difficult to visit a doctor) inspired me to make this project for helping needy people.

What it does

It can predict disease with great accuracy so that people can know what exactly happened and what to do.

How we built it

I have used one of the most successful and popular library XGBoost to train a ML classifier on a huge dataset of 4500+ datapoints which can predict a total of 41 diseases(including Asthma, Chicken Pox, Malaria, Pneumonia etc.). I created the website using Streamlit and showcased the ML model in that website.

Challenges we ran into

Handling such a large data with 4500+ rows and 120+ columns, visualizing the data, cleaning and reducing the features was really a challenge.

Accomplishments that we're proud of

We have achieved a significantly high accuracy in our classifier model with such a huge dataset which I consider as a milestone in my journey of Machine Learning.

What we learned

I gained some valuable experience of handling a huge dataset and the importance of data visualization in creating a successful ML model with large data.

What's next for DoctorAI

I will improve my Doctor by adding some more features soon which will include giving treatment for specific disease and symptoms, prescribing some easily available over the counter medicines and also I am trying to add the feature of live tracking nearby hospitals and pathology labs so that people can easily contact those for emergency situation and doing various tests and checkups.

Built With

• joblib
• numpy
• pandas
• python
• scikit-learn
• streamlit
• xgboost