input: training data set $D={\left\{\left(x_i,y_i\right)\right\}}_{i=1}^m$

output: Integrated classifier H

1. Learning base classifier

For $t=1,2,\cdots ,T$

Learn the base classifier $h_t$ on the data set D

2. Construct a new data set $D_h$

For $i=1,2,\cdots ,m$

$D_h=\left\{\left({x^{\prime }}_i,y_i\right)\right\}$ , in ${x^{\prime }}_i=\left\{h_1\left(x_i\right),\cdots h_T\left(x_i\right)\right\},i=1,2,\cdots ,m$

3. Learn a meta-classifier, H on the data set $D_h$

4. Return H