Algorithm 1: Adaboost classifier—pseudo code

Input: Let D be the dataset that includes {(a1,b1), (a2,b2), ….. (am, bm)};

Let λ be the learning (base) algorithm

Let T be the total No. of learning rounds.

Process:

D1(i) = 1/m

for time = 1, …, T;

ht = λ (D, Dt); weak learner is trained with Distribution Dt

$\in$t = $PrP{r}_{i~D_t}$[ $h_t$(ai$\ne b$i)]; Error measure (entropy)

${\propto }_t$ = $\frac{1}{2}$ ln ($\frac{1-\in \mathrm{t}}{\in \mathrm{t}});\% determine the weigth of$ ht

Dt+1(i) = $\frac{D_t\left(i\right)}{Z_t}$ ∗$\{\begin{array}{c}\exp \left(-{\propto }_t\right) if {\mathrm{h}}_{\mathrm{t}}\left({\mathrm{a}}_{\mathrm{i}}\right)=b_i\\ \exp \left({\propto }_t\right) if {\mathrm{h}}_{\mathrm{t}}\left({\mathrm{a}}_{\mathrm{i}}\right)\ne b_i\end{array}$

= $\frac{D_t\left(i\right)\exp (-{\propto }_t{y}_t{\mathrm{h}}_{\mathrm{t}}\left({\mathrm{a}}_{\mathrm{i}}\right)}{Z_t}$

Outcome: H(a) = sign (${{\displaystyle \sum }}_{t=1}^T{\propto }_t{h}_t\left(b\right)$