Premium Project

Clustering Heart Disease Patient Data

Experiment with clustering algorithms to help doctors inform treatment for heart disease patients.

Start Project
  • 10 tasks
  • 789 participants
  • 1,500 XP

Project Description

Doctors frequently study former cases in order to learn ways to best treat their patients. A patient who has a similar health history or symptoms to a previous patient could benefit from undergoing the same treatment. This project investigates whether doctors might be able to group together patients to target treatments using some common unsupervised learning techniques.

This project lets you apply clustering using the k-means and hierarchical clustering algorithms. We recommend that you take the following courses before starting this project: Data Visualization with ggplot2 (Part 1) and Unsupervised Learning in R.

The dataset for this project contains characteristics and measures of patients diagnosed with heart disease. It can be found here.

Project Tasks

  • 1Targeting treatment for heart disease patients
  • 2Quantifying patient differences
  • 3Let's start grouping patients
  • 4Another round of k-means
  • 5Comparing patient clusters
  • 6Hierarchical clustering: another clustering approach
  • 7Hierarchical clustering round two
  • 8Comparing clustering results
  • 9Visualizing the cluster contents
  • 10Conclusion
Instructor Avatar
Megan Robertson

Data Scientist at Nike

Megan Robertson is a data scientist with a background in machine learning and Bayesian statistics. She earned a Master's of Statistical Science from Duke University and has multiple years of experience teaching math and statistics. She is interested in sports analytics and interned with the Charlotte Hornets while in graduate school. Megan currently works as a data scientist at Nike.

See More


  • R LogoR
  • Topics

    Data ManipulationData VisualizationMachine LearningImporting & Cleaning Data