Cluster analysis

keywords: ALFA, clustering, cluster

ALFA provides support for cluster analysis, where in, a dataset or a set of variables can be clustered into k groups in an unsupervised fashion. ALFA currently supports the following algorithms

1. K-means
2. Spectral clustering
3. Hierarchical clustering

These algorithms can be invoked as shown below:

k means on [dataset] using k = 2
run spectral clustering on the [dataset] to group into 5 clusters

The output from the cluster analysis commands is displayed as a heatmap comparing the ground truth (reference) and output clusters.