Open AccessDistributed Ensemble based Deep Learning architecture for Intrusion Detection against Cyber attacks
Author(s) -
M. Kavitha,
K Elamukhil,
R Ajeeth,
R. Ashwin,
Vineeth N. Balasubramaniam
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1916/1/012080
Subject(s) - computer science , intrusion detection system , artificial intelligence , machine learning , server , deep learning , convolutional neural network , artificial neural network , ensemble learning , web server , classifier (uml) , network security , multilayer perceptron , perceptron , data mining , the internet , computer security , computer network , operating system
The increasing scale and importance of web contact around the Internet has increased the need for improved cyber security defence against cyber attacks. On modelling the machine learning based intrusion detection system, features of the attacks helps to discover, determine and identify unauthorized behaviour. Behaviour modelling has been used as training model to detect the evolving attacks to the servers or learning model has been constructed to build a training model to identify the intrusion in the network on basis of signature of the attacks. However machine learning model fails in handling attacks propagating on the applications with large scale or large dimensional data which further leads to high false alarm rate. In order to tackle those issues, a new distributed ensemble based deep learning architecture has been employed using Convolution Neural Network, Recurrent Neural Network and Multilayer Perceptron towards intrusion detection on cyber attacks on the web servers. Convolution Neural Network, Recurrent Neural Network and Multilayer Perceptron has been modelled as training model to detect the intrusion, trained model will generate best model which acts as classifier or prediction model for cyber attacks. The particular model will be employed further to detect the intrusion propagating in the web server. The proposed ensemble based deep learning architecture is compared to state-of-the-art approaches on performance measures such as precision, recall, and f measure on true positive, false positive, true negative, and false negative computations using the KDD CUP 99 dataset.