*3.4. Performance Measurements*

The statistical analysis included the calculation of the mean square error (*MSE*), Pearson's correlation coefficient (*R*), and the root-mean-square error (*RMSE*) to test the proposed algorithms' efficiency in detecting Android malware. The equations of these parameters are presented below:

$$MSE = \frac{1}{n} \sum\_{i=1}^{n} \left( y\_{i,exp} - y\_{i,\text{pred}} \right)^2 \tag{14}$$

$$RMSE = \sqrt{\sum\_{i=1}^{n} \frac{\left(y\_{i, \text{exp}} - y\_{i, \text{pred}}\right)^2}{n}} \tag{15}$$

$$R^{\diamondsuit}\_{\%} = \frac{n\left(\sum\_{i=1}^{n} y\_{i,\!xp} \times y\_{i,\!pred}\right) - \left(\sum\_{i=1}^{n} y\_{i,\!xp}\right)\left(\sum\_{i=1}^{n} y\_{i,\!pri}\right)}{\sqrt{\left[n\left(\sum\_{i=1}^{n} y\_{i,\!xp}\right)^2 - \left(\sum\_{i=1}^{n} y\_{i,\!xp}\right)^2\right]\left[n\left(\sum\_{i=1}^{n} y\_{i,\!pri}\right)^2 - \left(\sum\_{i=1}^{n} y\_{i,\!pri}\right)^2\right]}} \times 100\tag{16}$$

$$R^2 \ln 1 - \frac{\sum\_{i=1}^n \left( y\_{i,\exp} - y\_{i,\text{ pred}} \right)^2}{\sum\_{i=1}^n \left( y\_{i,\exp} - y\_{\text{avg\\_exp}} \right)^2} \tag{17}$$

$$Accuracy = \frac{TP + TN}{TP + FP + FN + TN} \times 100\% \tag{18}$$

$$Specificity = \frac{TN}{TN + FP} \times 100\% \tag{19}$$

$$Sensitivity = \frac{TP}{TP + FN} \times 100\% \tag{20}$$

$$\text{Precision} = \frac{TP}{TP + FP} \times 100\% \tag{21}$$

$$\text{Fscore} = \frac{2 \ast precision \ast \text{Sensitivity}}{precision + \text{Sensitivity}} \times 100\% \tag{22}$$

where *yi*,*exp* is the experimental value of the data point *i*, *yi*,*pre<sup>d</sup>* is the predicted value of the data point i, *yavg*,*exp* is the average of the experimental values, *R* is Pearson's correlation coefficient, *yi*,*exp* are the Android network packets of the input data i, *yi*,*class* are the classes of Android malware and normal input data i, *n* is the total number of the input data, the true positive (*TP*) represents the total number of samples that are successfully classified as positive sentiment, false positive (*FP*) is the total number of samples that are incorrectly classified as negative sentiments, true negative (*TN*) denotes the total number of samples that are successfully classified as negative sentiment, and false negative (*FN*) represents the total number of samples that are incorrectly classified as positive sentiments.
