11. Introduction to Unsupervised Learning Algorithms

Topics, Class Notes and Code Assignments and Side Readings
  • Concepts of unsupervised learning algorithms
  • Clustering
    • k-means clustering
    • Hierarchical clustering
  • Dimension reductions - principle component analysis (PCA)
  • Overview of anomaly detection algorithms
    • Types of anomaly detection methods
    • Supervised and unsupervised anomaly detection
  • Notes and code
  • Optional Reading Assignments
    • A survey of similarity measures in clustering PDF [link]
  • Written Assignment:Project #4 - Unsupervised Learning Algorithm -Part I
    • Pick a data set with a sample between 500 and 2000.
    • The data set should have at least 6 numerical variables (ideally to have some correlation).
    • Perform some EDA and potential feature engineering as usual.
    • Perform PCA and clustering analysis with this data.