In undersampling, we remove random amount of data from major ones to be similarly equal to the minor one.

One remedy for Handling Imbalanced Dataset; opposite of Oversampling.