Covid-19 Checker | Predictor

• 

  Homepage asking for user detail, so if the user is predicted to be positive then he/she can be contacted for the medical test.

• 

  Forms for medical conditions which can be filled up by the users, which then can be fed to the model to predict.

• 

  XGboost model for training

• 

  Accuracy can out to be 100% after a lot of efforts.

Inspiration

The COVID-19 virus has changed the lives of all. We all had faced a lot of losses of everything. There is no vaccine, the only way we can keep ourselves safe is by taking precautions. So, we made a ML model to predict the possibility of the infection of COVID-19.

What it does

Our model asks users about the important symptoms declared by the WHO in their guidelines and based on the response, it predicts the possibility. It also asks for whether the person has came in contact of COVID positive patient or not.

How we built it

We used XGBoost model to train the dataset available on the kaggle after preprocessing. After tuning all the parameters, we got the accuracy of 100%. For the frontend we used html, css, and javascript and backend was developed on Django.

Challenges I ran into

Major challenges was in the training the model and fine tuning the parameters as initially the model was not predicting anything; giving accuracy of 23% and then finally we got rid of it.

Accomplishments that I'm proud of

We altleast tried to complete the project, and done everything by our own, that makes me feel proud.

What I learned

I learned about how to fine tune the parameters and the syntaxes error the backend part and a lot more in the frontend. It was really a very tedious job.

What's next for Covid-Checker

More fine tuning, training on more data using deep learning, making it available for android, as a weapon against COVID-19.

Built With

• css
• django
• docker
• html
• javascript
• machine-learning
• python
• restframework