C

Covikit - The online Covid-19 test kit

This web application provides a platform for Online COVID-19 Testing. It is based on Artificial Neural Network and Exploratory Data Analysis. Built under Python & web designing was done on Streamlit.

The problem Covikit - The online Covid-19 test kit solves

COVID-19 Test can be done totally online, without physical interractions and free of cost.
Built on real-time Covid report dataset of 100 thousand people around the world. Employees Deep Learning and Neural Networks.
Makes the Covid test safer as it is totally virtual.
After the test, it awares the patient on the next step. If they test positive, it prescribes them the next crucial steps.
If they test negative, it suggests best precautions to avoid covid19.
Allows the patient to live "Quarantine & chill" all safe with it's inbuilt dark mode theme.

Challenges we ran into

  1. Huge dataset processing - preprocessing and cleaning the data helped.
  2. Factor dependencies - Characteristic factors were focused.
  3. Test Prediction Accuracy - 8 ML and 2 DL Models were employed and the best one was engineered for model building.

Discussion