Horse Colic Dataset Project README

Summary and Motivation

The purpose of this project is to explore the Horse Colic dataset from UC Irvine and use it to answer questions with real-world application in horse health and care, such as:

1. What characteristics are most associated with cases in which the colic needed surgery to be treated?
2. What characteristics of the horse's condition were most associated with a "lived" outcome (as opposed to "died" or "was euthanized")?
3. What were some of the ranges of the various health indicator parameters commonly measured by horse owners (such as rectal temperature, heart rate, respiratory rate, capillary refill time, mucous membranes and pain level)?

The motivation for this project is to satisfy requirements for Term 2 of the Udacity Data Scientist Nanodegree, as well as to practice and demonstrate fundamental data science skills and the data science process.

Libraries Used

• matplotlib.pyplot for visualizations
• numpy
• pandas

Files Included

• Horse_Colic_Project.ipynb -- Jupyter notebook containing write-up of project
• Horse_Colic_Project.html -- HTML version of Jupyter notebook containing write-up of project
• horse-colic-dataset/datadict.txt -- Text document summarizing dataset
• horse-colic-dataset/horse.csv -- CSV file containing dataset
• README.md -- This README
• BlogPost.md -- Blog Post containing write-up of the project for a non-technical audience
• horse-grazing.jpeg -- Horse stock photo for Blog Post
• tree.dot -- Visualization of Decision Tree classifier
• tree.png -- Visualization of Decision Tree classifier
• tree2.dot -- Visualization of Decision Tree classifier
• tree2.png -- Visualization of Decision Tree classifier
• vital_signs.png -- Plot of categorical vital signs
• vital_signs_num.png -- Plot of numerical vital signs

Summary of Results

The results of this analysis are summarized in a Blog Post, which can be found in this GitHub repo under BlogPost.md.

Acknowledgements

UCI for Horse Colic Dataset.