Skripsi

OPTIMALISASI PERFORMA ALGORITMA CONVOLUTIONAL NEURAL NETWORK (CNN) DALAM PROSES KLASIFIKASI MALWARE BOTNET PADA JARINGAN INTERNET OF THINGS

---

Penilaian

---

One of the threats to the internet network is botnet (robot network). although there are many methods used to detect botnet, but there are still less accurate. this can be seen from the results of accuracy, precision and others that differ greatly due to imbalanced datasets. This research examines the classification of botnet attacks and builds the best and accurate CNN model by optimizing the CICIDS-17 dataset consisting of 97718 BENIGN data and 128027 DDoS data by applying undersampling techniques and AlexNet and LeNet architectures in the Convolutional Neural Network method. After being optimized, AlexNet architecture gets an accuracy of 99.97% from 99.94% and the loss value decreases from 0.49% to 0.11%. while Lenet architecture the accuracy increases from 99.88% to 99.93% and the loss value decreases from 0.40% to 0.24%. Based on the accuracy graph, both models are neither overfitting nor underfitting. Meanwhile, based on the confusion matrix, it can be seen that the model is able to classify botnets quite well. Keywords : Convolutional Neural Network, CICIDS-17, Botnet IoT, AlexNet, LeNet

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2307006199	T129898	T1298982023	Central Library (Referens)	Available

Detail Information

Series Title

-

Call Number

T1298982023

Publisher

Indralaya : Prodi Sistem Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya., 2023

Collation

xvii, 95 hlm.; ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

005.707

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Sistem Komputer
Data dalam sistem-sistem komputer

Specific Detail Info

-

Statement of Responsibility

MURZ

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment