Data only comes with input X, but not label output label Y.
Algorithm has to find some structure in this data.
Clustering: Group similar data points together
Anomaly Detection: Finds unusual data points
Dimensionality Reduction: Compress data using fewer numbers