Bagging is a technique used in random forests to:

$$Prune the decision tree to reduce overfitting

$$Determine the optimal number of clusters in K$-$means clustering

$$Boost the performance of individual decision trees

$$Reduce the computational cost of training the ensemble