Skripsi
OPTIMALISASI TEKNIK KLASIFIKASI SERANGAN BOTNET DENGAN MENGGUNAKAN SELEKSI FITUR HASHMAP DAN K-NEAREST NEIGHBOR
A botnet consists of a group of interconnected software programs that communicate with each other via the internet to perform designated functions. This software is programmed to operate automatically in the network. A botnet driver, also known as a bot master, remotely controls each computer that is part of a botnet network. Therefore, it can be concluded that if a computer is infected with a botnet, after connecting to the network, it will execute commands issued by the bot master. K-Nearest Neighbor(K-NN) is a classification method that uses the majority of categories in K-NN to determine the category of new data or data testing. In K-NN, k objects clossest (similiar) to the new data objext are searched in the training data. Results on the Confusion Matrix Algorithm 50:50 Precision Accuracy Recall F1-Score KNN 99.80% 99.77% 99.84% 99/80%. The results of the value of 99.80%. Furthermore, the results and analyss of the scenarios used in the Hashmap Hashmap ROC curve TPR FPR KNN 99.75% 0.27%. The results of the average value on the ROC curve using the hashmap selection feature using the KNN method with a TPR value of 99.75% Results and analys of the scenarios used in the hashmap selection feature used with the confusion matrix Results in the Confusion Matrix Algorithm 70:30 F1 Recall Precision Accuracy - KNN score 99.71% 99.67% 99.75% 99.71%. The average value of the hashmap selection feature uses the KNN method with a value 99.71%. Futhermore, the results and analys of the scenario used in the ROC on the TPR FPR KNN Hashmap Algorithm are 99.75% 0.32%. The results of the average value of 99.75%. Results on the Confusion Matrix Algorithm 90:10 Presision Accuracy Recall F1-Score KNNN 99.79& 99.75% 99.83% 99.97%. The results of visualization of parallel coordinates between benign and botnet data for blue patterns show binign and green colors using Hashmap to perfom feature selection where there are 52 features, the features used, the faster the training and the features that have been selected are relevant features
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2307005624 | T128575 | T1285752022 | Central Library (Referens) | Available |
No other version available