Table 1. Algorithm 1. Bagging.

1: Let T be the set of n training examples (x i, y i), i ∈ 1, 2, ⋯, n.

2: B is the number of base learners and L the base learning algorithm.

3: for(i = 0; i < B; i + +){

4:  Create a bootstrapped training set T i of size n by sampling with replacement.

5:  Learn a specific base learner L i(x, y) on T i by using L.

6: }

7: The final learning algorithm C is the ensemble of all base learners {L i} and a test example x* is classified by using a simple majority voting method: $$\begin{array}{c}\hfill y^{*}=arg\underset{y}{max}\sum _{L_i\in C}{L}_i({\mathrm{x}}^{*},y)\end{array}$$