Respiratory-Mortality

Predict respiratory patient mortality in ICU units using the MIMIC III database. PostgreSQL was used for feature engineering. AWS EC2 instance with Rstudio used to build a global dataset, data cleaning, and model building. Tableau and R used for visualization.

Name		Name	Last commit message	Last commit date
Latest commit History 46 Commits
R_Code		R_Code
sql_code		sql_code
MEDIAN LOS BY DIAGNOSIS.JPG		MEDIAN LOS BY DIAGNOSIS.JPG
README.md		README.md
compare different.png		compare different.png
comprae all models.png		comprae all models.png
correlation matrix.JPG		correlation matrix.JPG
full model comparison.png		full model comparison.png
histograms.JPG		histograms.JPG
mimic entity relationship diagram.png		mimic entity relationship diagram.png
model_accuracies.xlsx		model_accuracies.xlsx
mortality by admission_location.JPG		mortality by admission_location.JPG
mortality by age.JPG		mortality by age.JPG
target by admission type where patient died.JPG		target by admission type where patient died.JPG
target count by admission location.JPG		target count by admission location.JPG
unbalanced target.JPG		unbalanced target.JPG
variable worth.JPG		variable worth.JPG
vent duration by admission location and hospital expire flag.JPG		vent duration by admission location and hospital expire flag.JPG

AGARNER18/Respiratory-Mortality

Folders and files

Latest commit

History

Repository files navigation

Respiratory-Mortality

About

Topics

Resources

Stars

Watchers

Forks

Languages