zgk2003 / covid19_prediction

This is a project to determine COVID-19 infection in advance from available symptoms.
1 stars 0 forks source link

Peer Review #2

Open az-cheng opened 3 years ago

az-cheng commented 3 years ago

The project is about using COVID-19 data to help guide decision making. More specifically, they want to identify patients from the symptoms they display. They use data collected globally from Facebook users.

I like about:

  1. the size of the dataset and the abundance of information it contains
  2. how you want to develop a "standardized, global way" to measure covid
  3. you're already thinking about models that you may use to fit the data

I think it can be improved by:

  1. taking into account bias of only collecting data from Facebook users
  2. the data only spans a short period of time from April 6 to May 1, 2020, how would you respond to symptoms changing as variants develop?
  3. similarly, how would you address the different symptoms displayed in regions dominating by different variants?