*3.4. Machine Learning Algorithms*

After the selection of features, classification is performed. SVM (Support Vector Machine) is a classifier for binary classification of data. The hyperplane is used to solve the learning problem in SVM. A robust method with different kernel values is considered one of the best classifiers for classification [37]. RF (Random Forest) utilized various trees to predict. It is being utilized by different research areas of research with remarkable results. RF produces high classification accuracy with an even dataset with a large number of features. It handles unbalanced data by accessing important features. Whereas GBDT (Gradient Boosting Decision Tree) is selected due to its property of selecting fewer parameters as compared to the other classification algorithms. In existing research, in machine learning, GBDT shows tremendous results. It is based on the CART algorithm. GBDT merges the concept of regression and boosting tree and intends the use of residual gradient to optimize the assimilation process of regression tree [38]. ANN (Artificial Neural Network) is a popular classification technique utilized in different areas of research like agriculture, medical, security, education, business, art, etc. It is very easy to use and can manage complex data [39]. Moreover, the performance of the proposed approach presented in this paper is evaluated through accuracy, precision, recall, and f-measure, whereas accuracy is defined as the predicted observations over a total number of observations [40–42]. Precision is the fraction of the recovered instances that belong to the target class, whereas F-measure is the harmonic mean of precision and recall. Equations (3)–(6) presents the formula of evaluation parameters, whereas *TP*, *FN*, and *FP* stand for true positive, false negative, and false-positive respectively.

$$\text{Accuracy} = \frac{TP + FN}{TP + FN + FP + FN} \tag{3}$$

$$\text{Precision} = \frac{TP}{TP + FP} \tag{4}$$

$$\text{Recall} = \frac{TP}{TP + FN} \tag{5}$$

$$\text{F}-\text{Measure} = \frac{\text{2(Precision} \times \text{Recall)}}{\text{Precision} + \text{Recall}} \tag{6}$$

whereas Table 3 presents the Parameters of classification algorithm utilized in proposed work.


**Table 3.** Parameters of classification algorithm utilized in proposed work.
