A Strong Machine Learning Classifier and Decision Stumps Based Hybrid AdaBoost Classification Algorithm for Cognitive Radios

. 2019 Nov 20;19(23):5077. doi: 10.3390/s19235077

Algorithm 1 A Strong Machine Learning Classifier and Decision Stumps Based Hybrid Adaboost Classification Algorithm.

Require:

\bar{X}, {c^{(1)}, c^{(2)}, \dots, c^{(L)}}

X^{*}

T

(number of sub-classifiers).
Ensure:

{\hat{c}}^{(m)}, m \in 1, 2, \dots, M

1:
Initialize sample weight $D_{1} (l) = \frac{1}{L}$
2:
$ε_{1} = \sum_{l = 1}^{L} D_{1} (l) I (h_{1} (x^{(l)}) \neq c^{(l)})$
3:
$α_{1} = \frac{1}{2} ln (\frac{1 - ε_{1}}{ε_{1}})$
4:
$D_{2} (l) = \frac{D_{1} (l) exp (- α_{1} c^{(l)} h_{1} (x^{(l)}))}{Z_{1}}$
5:
for $t = 2 t o T$ do
6:
$ε_{t} = \sum_{l = 1}^{L} D_{t} (l) I (h_{t} (x^{(l)}) \neq c^{(l)})$
7:
$α_{t} = \frac{1}{2} ln (\frac{1 - ε_{t}}{ε_{t}})$
8:
$D_{t + 1} (l) = \frac{D_{t} (l) exp (- α_{t} c^{(l)} h_{t} (x^{(l)}))}{Z_{t}}$
9:
end for
10:
Predict the class of the m-th test sample ${\hat{c}}^{(m)} = s i g n (\sum_{t = 1}^{T} α_{t} h_{t} (x^{* (m)}))$