Abstract Botnet threat detection has been a focus of continuing study. Botnet identification using flow-based features has been successfully accomplished using machine learning (ML) approaches.Flow-based features' main drawbacks are their significant processing expense and partial capture of network communication patterns.This research propose novel technique in BotNet prediction among the authorized social media users by machine learning algorithm feature analysis. Information has been gathered here from users of social media platforms also it has been filtered based on unusual activities. Then this filtered data features has been extracted and classified using KNN-Xception architecture where the malicious activity. An assessment of the experimental data has been done with regards of detection accuracy, RMSE, malicious activity rate, recall, mAP. The suggested method accomplished detection accuracy of 96%, RMSE of 61%, malicious activity rate of 39%, recall of 59%, mAP of 61%.
Alan : Mühendislik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|