Inspiration
We got inspired to be part of this challenge so that we could help build a solution to our community. So many people here are not aware of how they can get tested to know their status for the COVID-19 disease.
What it does
We have submitted an app that provides for self testing based on WHO guidelines for the COVID19 disease using the pre-exposure checks.
How we built it
We used a python flask framework to build the main application for data collection via web form and dashboard. Then we built a classification model using scikit-learn that was able to predict whether a test is required or not based on features given. For training the model, we asked a medical colleague to help us with labelling the data on recommendations based on the test data submitted. Then the model was trained using these labels and finally we wrote our report. We used Africa's Talking USSD simulator to build the USSD front-end and test it as well.
Challenges we ran into
- Insufficient training data
- Time was very short.
- Internet connectivity challenges due to working at home advisories.
Accomplishments that we're proud of
- Designed a working application that can be used to solve COVID-19 triaging and tracking challenges.
- Levelled up on our collaboration skills
What we learned
- Learning process - we have learnt so much from this hackathon.
- Embedding the machine learning model into a live system.
What's next for A MODEL TO DECREASE TRANSMISSION OF COVID19 IN KENYA
Polishing and implentation of the model and then rolling it out for public use to aid in the fight against COVID-19.
Log in or sign up for Devpost to join the conversation.