Analisis Cloud Forensic dalam Mendeteksi Anomali Menggunakan Metode Long Short-Term Memory: Studi Kasus pada Data Center D-Net Surabaya

Alwan, Muhammad (2024) Analisis Cloud Forensic dalam Mendeteksi Anomali Menggunakan Metode Long Short-Term Memory: Studi Kasus pada Data Center D-Net Surabaya. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5027201019-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Di era digital yang terus berkembang, cloud computing memainkan peran krusial dalam menyediakan layanan yang efisien dan skalabel. Namun, dengan peningkatan pemanfaatan cloud, keamanan menjadi perhatian utama, khususnya dalam deteksi dan penanganan anomali. Cloud forensics menjadi penting dalam mengidentifikasi aktivitas yang mencurigakan atau tidak wajar yang menunjukkan adanya potensi ancaman. Penerapan teknik pembelajaran mesin seperti Long Short-Term Memory (LSTM) menjanjikan kemajuan dalam memahami dan mendeteksi anomali, mengingat kemampuannya dalam memodelkan dependensi waktu yang kompleks dalam data berurutan.
Penelitian ini mengusulkan penerapan LSTM untuk mengevaluasi dan menganalisis anomali dalam data cloud di Data Center D-Net Surabaya. LSTM, dengan kemampuan memori jangka panjang dan pendeknya, diharapkan dapat mengatasi tantangan dalam mendeteksi perilaku tidak biasa dalam sejumlah besar data cloud. Metode ini melibatkan pelatihan model LSTM dengan data historis untuk mengenali pola aktivitas normal dan anomali, memungkinkan deteksi anomali yang lebih akurat dan responsif terhadap perubahan pola data.
Hasil penelitian menunjukkan bahwa model unsupervised biderectional LSTM yang telah dioptimiasi dengan hyperparameter dan XGBoost mampu mendeteksi anomali pada data cloud lebih baik dibandingkan model unsupervised lain yaitu K-Means Clustering. Model LSTM mencapai nilai MSE 0,2690, jauh lebih rendah dibandingkan metode K-means dengan MSE 0,7572. Hal ini menunjukkan bahwa model LSTM dapat diimplementasikan untuk meningkatkan keamanan arsitektur cloud computing.
====================================================================================================
In the ever-evolving digital era, cloud computing plays a crucial role in providing efficient and scalable services. However, with the increase in cloud utilization, security is a major concern, particularly in anomaly detection and handling. Cloud forensics, particularly anomaly detection, is important in identifying suspicious or unusual activity, which may indicate a potential threat or security breach. The application of machine learning techniques such as Long Short-Term Memory (LSTM) promises progress in understanding and detecting anomalies, given its ability to model complex time dependencies in sequential data.
This research proposes the application of LSTM to identify and analyze anomalies in cloud data at the D-Net Surabaya Data Center. LSTM, with its long and short-term memory capabilities, is expected to overcome the challenge of detecting unusual behavior in a large amount of cloud data. This method involves training the LSTM model with historical data to recognize normal and anomalous activity patterns, enabling more accurate anomaly detection and responsiveness to changes in data patterns.
The research results show that the unsupervised bidirectional LSTM model that has been optimised with hyperparameters and XGBoost can detect anomalies in cloud data better than other unsupervised models, such as K-Means Clustering. The LSTM model achieves an MSE of 0.2690, which is much lower than that of the K-means method with an MSE of 0.7572. This shows that the LSTM model can be implemented to improve the security of the cloud computing architecture.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Cloud Forensic, Deep Learning, Deteksi Anomali, Long Short-Term Memory, Anomaly Detection, Cloud Forensic, Deep Learning, Long Short-Term Memory
Subjects:	T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Alwan
Date Deposited:	02 Aug 2024 03:26
Last Modified:	02 Aug 2024 03:27
URI:	http://repository.its.ac.id/id/eprint/112125

Actions (login required)

View Item