Academia.eduAcademia.edu

Figure 2 – uploaded by Hajara Musa

See full PDFdownloadDownload figure
phishing. The mathematical equations of the performance metrics are given below respectively.

Figure 2 phishing. The mathematical equations of the performance metrics are given below respectively.

Related Figures (4)

Figurel .0: Research framework Figure 1.0 illustrate the stages followed in this research where some selected papers were reviewed in order to determine the research gap and as such the research problem was formulated; special attention was given to feature selection, phishing website detection, classification, and clustering. Others includes; random forest, probabilistic neutral network, k-nearest neighbor, naive bayes and extreme gradient boosted tree techniques as these are the constituents of this research work. Special attention was also given to data pre- processing being a crucial stage in classification task. It is interested to know that, most researchers in phishing detection make use of datasets constructed by them. However, with such type of datasets, it is difficult to evaluate and compare the performance of a model with other models from the literature since the datasets used were not publicly available for others to use and confirm their results, therefore such results cannot be generalized [7].
Figurel .0: Research framework Figure 1.0 illustrate the stages followed in this research where some selected papers were reviewed in order to determine the research gap and as such the research problem was formulated; special attention was given to feature selection, phishing website detection, classification, and clustering. Others includes; random forest, probabilistic neutral network, k-nearest neighbor, naive bayes and extreme gradient boosted tree techniques as these are the constituents of this research work. Special attention was also given to data pre- processing being a crucial stage in classification task. It is interested to know that, most researchers in phishing detection make use of datasets constructed by them. However, with such type of datasets, it is difficult to evaluate and compare the performance of a model with other models from the literature since the datasets used were not publicly available for others to use and confirm their results, therefore such results cannot be generalized [7].
Table 1.0 shows the confusion matrix in which TP (True Positive) is a case where a model correctly predicts a website as phishing, TN (True Negative) is a case where a website is wrongly classified as benign. FP (False Positive) is a case where a website is wrongly classified as phishing and lastly FN (False negative) is when the model wrongly classified a website as benign while it is actually To evaluate and compare the performance of our proposed model with other models from the literature, the following evaluation metrics were employed; accuracy (ACC) , precision (Prec), recall (Rec), mathew correlation coefficient (MCC), and f-score. ACC measures the ratio of websites which are correctly predicted. Prec measures the fraction of websites correctly predicted as phishing. Rec metric measures the fraction of phishing websites identified by the model.
Table 1.0 shows the confusion matrix in which TP (True Positive) is a case where a model correctly predicts a website as phishing, TN (True Negative) is a case where a website is wrongly classified as benign. FP (False Positive) is a case where a website is wrongly classified as phishing and lastly FN (False negative) is when the model wrongly classified a website as benign while it is actually To evaluate and compare the performance of our proposed model with other models from the literature, the following evaluation metrics were employed; accuracy (ACC) , precision (Prec), recall (Rec), mathew correlation coefficient (MCC), and f-score. ACC measures the ratio of websites which are correctly predicted. Prec measures the fraction of websites correctly predicted as phishing. Rec metric measures the fraction of phishing websites identified by the model.
Table 2.0 Comparison of Results The predictive performance of phishing website detection model using XGBOOST
Table 2.0 Comparison of Results The predictive performance of phishing website detection model using XGBOOST
Figure 2.0 Accuracy Chart Representation of Models Comparison algorithms was optimized; thereby bringing the models much closer to accurate measurement of the performance metric than other models in which their models performance is less than our own in case of precision, recall, f.score, MCC and Accuracy. As we can see from Table 2.0 PNN has 0.9640, 0.9789, 0.9714, 0.9350 and 0.9679 accuracies respectively and RF has 0.9430, 0.9796, 0.9611, 0.9128, and 0.9566 accuracies respectively while XGBOOST returned the highest accuracies of 0.9730, 0.9801, 0.9724, 0.9449, and 0.9729 respectively. Therefore the proposed model has the highest accuracy compared with the rivals models. The result can be represented in graphical form for better analysis and understanding.
Figure 2.0 Accuracy Chart Representation of Models Comparison algorithms was optimized; thereby bringing the models much closer to accurate measurement of the performance metric than other models in which their models performance is less than our own in case of precision, recall, f.score, MCC and Accuracy. As we can see from Table 2.0 PNN has 0.9640, 0.9789, 0.9714, 0.9350 and 0.9679 accuracies respectively and RF has 0.9430, 0.9796, 0.9611, 0.9128, and 0.9566 accuracies respectively while XGBOOST returned the highest accuracies of 0.9730, 0.9801, 0.9724, 0.9449, and 0.9729 respectively. Therefore the proposed model has the highest accuracy compared with the rivals models. The result can be represented in graphical form for better analysis and understanding.
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved