-
The goal of this repo is to provide a realistic setting for a machine learning application. The UCI dataset includes over 50 features representing patient and hospital outcomes.
-
The aim is to predict which patients will need hospital readmission.
-
This tutorial covers most steps in the ML pipeline:
-
1 Understanding the dataset
-
2 Data Assembling and Preprocessing
-
3 Machine Learning Pipeline
-
4 Alternative Pipeline
-
5 Model Interpretation
-
6 Limitations and potential Solutions
-
To run the notebook locally just install all required packages.
pip install -r requirements.txt