# APELID

**Repository Path**: xbystudy/APELID

## Basic Information

- **Project Name**: APELID
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-07-24
- **Last Updated**: 2024-07-24

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# APELID: Enhancing Real-Time Intrusion Detection with Augmented WGAN and Parallel Ensemble Learning


Perform an approach that combines a rule-based and ensemble learning model for the intrusion detection system.

For the Manuscript "APELID: Enhancing Real-Time Intrusion Detection with Augmented WGAN and Parallel Ensemble Learning"!

In order to validate and evaluate the effectiveness of the PELID method, we use two sub-datasets, CSE-CIC-IDS2018  and NSL-KDD.

1. The first dataset (DS1) includes CSE-CIC-IDS2018. The goal of using this dataset is to verify PELID's ability to detect network attacks by analyzing the network traffic correctly. These datasets were also selected based on assessing the SOTA methods.

2. The second dataset (DS2): We use the classic NSL-KDD and the up-to-date CSE-CIC-IDS2018 as benchmark datasets. For our experiment, within the dataset exists, we divided it into four different classes of attacks: (1) Denial of Service (DoS), which tries to shut down the traffic flow to and from the target; this is the most common attack in the dataset; (2) Probe, this kind of attacks, tries to get information from a network the goal of this attacks is to act like a thief and steal important information; (3) User to Root (U2R), such as privilege escalation attacks, with a normal user account and tries to gain access to the system or network, as a super-user(root); Remote to Local(R2L), tries to gain access to a remote machine, an attacker do not have local access to the system/network and tries to hack into the network. Thus, our goal when using this data set is to objectively compare the efficiency of the PELID method with other studies using the same dataset.

From these two sub-datasets, we constitute a dataset for testing. The dataset contains two parts: training and testing at the ratio of 7:3.

For full access to the source code and datasets, please download them by following the link: 
https://drive.google.com/drive/folders/1wc7BwyjjEI7c6IBKg5Phv7HRYmLxqpnX?usp=share_link

With the contributions of the authors:

Hoang V. Vo - Department of Information Systems, VNU University of Engineering and Technology, Hanoi 100000, Vietnam

Hanh P. Du - Department of Information Systems, VNU University of Engineering and Technology, Hanoi 100000, Vietnam

Hoa N. Nguyen - Corresponding author - hoa.nguyen@vnu.edu.vn - Department of Information Systems, VNU University of Engineering and Technology, Hanoi 100000, Vietnam