Mortality Status Detection Analysis

A project on a classification problem.

About the dataset:

The icu.csv dataset contains the following information about 20,000 patients in ICU:

• Patient_ID
• Mortality
• Age
• Race
• SBP
• DBP
• MAP
• Temperature
• Respiration

You can download the icu.csv file from the data folder.

Aim of the project:

To perform the following tasks:

• Do exploratory data analysis to identify variables that are significant predictors to predict the mortality status of patients.
• Perform classification and build a model to predict the mortality status of patients when values for the identified predictors are given as an input.
• Use regularization methods to address overfitting/underfitting issues in the model.
• Discuss the shortcomings of the model.
• Identify if there are any biases in the dataset and suggest how to overcome them.

Instructions to view the analysis:

1. Run the R markdown file:

• Download the ICU-MortalityStatusDetectionAnalysis.Rmd file.
• Set directory with the ICU-MortalityStatusDetectionAnalysis.Rmd file as the working directory.
• Download the icu.csv dataset within the data folder in the working directory.
• Run the ICU-MortalityStatusDetectionAnalysis.Rmd file.

2. Download and view the ICU-MortalityStatusDetectionAnalysis.html file.