16.3 Clustering Algorithms

Many clustering algorithms exist:

1. Ward’s hierarchical clustering
2. K-means clustering
3. Hierarchical clustering
4. Density-based clustering
5. Spectral clustering
6. Mean shift
7. Affinity propagation
8. Mixture model (latent profile analysis, latent class analysis)

The majority of class today will cover (1) Hierarchical and (2) K-Means Clustering.

16.3.0.1 Hierarchical Clustering

• We do not know in advance how many clusters we want
• Bottom-up approach (grouping similar observations together)
• End up with a tree-like visual representation of the observations called a dendrogram

16.3.0.2 K-means Clustering

K-means clustering:

• Seek to partition the observations into a pre-specified number of clusters
• Top-down approach