project 1

Experimental Study of PAGA and COVID-19 Patient Classification

This project aims at classifying the stage of COVID-19 patients (recovered, non-ICU, and ICU) according to the abundance of different types of cells in their blood samples using KNeighborsClassifier.

This project achieved the following:

  • Gathered blood samples from 20 recovered, 5 non-ICU, and 10 ICU COVID-19 patients.
  • Used python Scikit-Learn’s LASSO and LinearSVC functions to conduct dimension reduction on the samples and studied on the top 5 dominate cell types.
  • Used Python Scikit-Learn’s StratifiedKFold function to split the data into 5 training sets and 1 validation set and conducted supervised learning.

The GitHub repository of this project can be found at here, which includes both the code and the results.