Enhanced Learning Vector Quantization for Detecting Intrusions In IDS

Enhanced Learning Vector Quantization for Detecting Intrusions In IDS

Sandosh S. (Pondicherry Engineering College, Pondicherry, India), Govindasamy V. (Pondicherry Engineering College, Pondicherry, India) and Akila G. (Pondicherry Engineering College, Pondicherry, India)
Copyright: © 2020 |Pages: 16
DOI: 10.4018/IJWP.2020010105
OnDemand PDF Download:
List Price: $37.50
10% Discount:-$3.75


Nowadays, computer infrastructure attacks have become more challenging with computer network extension. The detection of intrusion is the most important process against threats. Several traditional methods are there, but still, there is an issue in detecting the errors related to security. In this article, the intrusion detection system is performed through various detection methods. The efficient proposed method is the Enhanced Learning Vector Quantization (ELVQ) algorithm for detecting the intrusions presented in network traffic. Also, the Pearson Correlation Coefficient Function (PCCF) for similarity determination is introduced. The proposed ELVQ classification achieves higher classification accuracy. In the end, risk factors are analyzed using Hidden Bernoulli Model (HBM). The proposed system is evaluated with KDD CUP99 dataset for efficient results. The evaluation results proved the proposed method performance with various measures and compared with various methods such as Random forest, Random tree, MLP, and Naïve Bayes.
Article Preview


The intrusion detection system (IDS) performs the most important role in information security and it can be identifying the various network attacks accurately. It is one of the benchmark methods to protect personal computer from many attacks. The network intrusion detection system (NIDS) is used for administrators to detect network threads in their organization. The Network Intrusion Detection Systems are presented in network points to monitor the traffic. Intrusion detection system works based on attacks characteristics, tracking of unauthorized access and malicious activities. For developing the internet and network technology, the intrusion detection system is mostly used for identifying the attacks. The various security technologies are information encryption; access control and prevention of intrusion which are used for protecting the network systems. But these systems still have some errors. There are various problems in intrusion detection based on various large numbers of false alerts.

To solve this issue, an enhanced intrusion detection system is proposed. It contains the data mining classification for false alerts handling in the intrusion detection system. Our proposed detection system is enhanced with the enhanced learning vector quantization (ELVQ) algorithm. The intrusion detection system risk assessment analysis is done by using hidden Bernoulli Model. This HBM model is hidden of the Markov model and it is generalized with the Bernoulli process.

Also, the Pearson Correlation Coefficient Function (PCCF) is used for making it is more efficient in anomaly detection. It’s mainly used for similarity detection. Initially, the data set is pre-processed with the Interquartile Range outlier detection technique in the proposed method. The whole process is implemented by efficient KDD-CUP 99 dataset. The proposed method evaluation results proved with KDD-CUP dataset and measured the proposed method performances. In this system, the performance for the detection of intrusions is improved. It shows various types of attacks and it is compared with the proposed technique for performance measures.

Figure 1.

Intrusion Detection System (IDS)


Figure 1 shows the basic concept of Intrusion Detection System (IDS). It defines the software application and it monitors the system activities then provides reports to management. It also has a network-based intrusion detection system and host-based network intrusion detection system. The IDS system aims to achieve malicious activities in various ways.




The main contribution of this work is

  • To improve the pre-processing of a dataset using Interquartile Range outlier detection (IRD)

  • To enhance the similarity detection the Pearson Correlation Co-efficient Function (PCCF) is used.

  • To estimate the classification outcomes the Enhanced Learning Vector Quantization (ELVQ) algorithm is used.

  • To improve the performance of intrusion detection system an efficient KDD-CUP 99 dataset is used.

Organization of Paper

First Section defines the importance of information security and detection system. Second Section contains various reviews of the intrusion detection system and errors in IDS. Third Section provides an explanation of the proposed Enhanced Learning Vector Quantization (ELVQ) method with KDD-CUP 99 dataset. Fourth Section defines the various performance measures of the proposed system. Final Section provides the conclusion of the proposed method.

Complete Article List

Search this Journal:
Volume 15: 1 Issue (2023): Forthcoming, Available for Pre-Order
Volume 14: 2 Issues (2022): 1 Released, 1 Forthcoming
Volume 13: 2 Issues (2021)
Volume 12: 2 Issues (2020)
Volume 11: 2 Issues (2019)
Volume 10: 2 Issues (2018)
Volume 9: 2 Issues (2017)
Volume 8: 1 Issue (2016)
Volume 7: 2 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing