Algorithm 1 CNN-based IoT attack detection

Require: Dataset $D=\{X,y\}$, learning rate ${\eta }_0$, batch size B, epochs E, folds K

Ensure: Trained model $\mathcal{M}$, results $\mathcal{R}$

1: Normalize features: $x^{\prime }={\displaystyle \frac{x-x_{min}}{x_{max}-x_{min}}},\forall x\in X$

2: Initialize model weights $W,b\sim \mathcal{N}(0,{\sigma }^2)$

3: Initialize result set $\mathcal{R}\leftarrow \varnothing$

4: $F\leftarrow$ Stratified K-fold split on $(X,y)$

5: for all $(X_{\mathrm{train}},y_{\mathrm{train}},X_{\mathrm{test}},y_{\mathrm{test}})\in F$ do

6:     $(X_{\mathrm{train}}^{\mathrm{res}},y_{\mathrm{train}}^{\mathrm{res}})\leftarrow \mathrm{SMOTE}(X_{\mathrm{train}},y_{\mathrm{train}})$

7:     for $e=1$ to E do

8:         Shuffle$(X_{\mathrm{train}}^{\mathrm{res}},y_{\mathrm{train}}^{\mathrm{res}})$

9:         for all $B_i=(X_B,y_B)\subset (X_{\mathrm{train}}^{\mathrm{res}},y_{\mathrm{train}}^{\mathrm{res}})$ do

10:            $Z=W\ast X_B+b$

11:            $\widehat{Z}={\displaystyle \frac{Z-\mu }{\sqrt{{\sigma }^2+ϵ}}}$

12:            $\tilde{Z}=\mathrm{Dropout}(\widehat{Z},p)$

13:            $\widehat{y}={\displaystyle \frac{e^{\tilde{Z}}}{\sum e^{\tilde{Z}}}}$

14:            $L=-{\sum }_{i=1}^N{y}_{i}log\left({\widehat{y}}_i\right)$

15:            $W\leftarrow W-\eta {\displaystyle \frac{\partial L}{\partial W}}, b\leftarrow b-\eta {\displaystyle \frac{\partial L}{\partial b}}$

16:         end for

17:         $\eta \leftarrow {\eta }_0\times {\lambda }^{{\displaystyle \frac{e}{E}}}$

18:         ${\mathcal{R}}_e\leftarrow \mathrm{Evaluate}(\mathcal{M},X_{\mathrm{test}},y_{\mathrm{test}})$

19:         if EarlyStopping(${\mathcal{R}}_e$) then

20:              Break

21:         end if

22:     end for

23:     Append ${\mathcal{R}}_e$ to $\mathcal{R}$

24: end for

25: return $\mathcal{M},\mathcal{R}$