ANALYSIS OF ARTIFICIAL NEURAL NETWORKS BASED INTRUSION DETECTION SYSTEM

  • Mohith Sai Krishna.Katakam , Komali.Devineni, Pavana.Kanagala, DR.N.Raghavendra sai

Abstract

A decent method to identify ill-conceived use is through checking unordinary client action. Techniques for interruption identification dependent close by coded rule sets or foreseeing directions on-line are laborous to construct or not truly solid. This paper proposes another method for applying neural systems to identify interruptions. In the proposed technique, rather than thinking about all highlights for interruption recognition and burning through up the time in investigating it, just the important component for the specific assault is chosen and interruption discovery is finished with assistance of managed learning Neural Network (NN). The element determination is finished with the assistance of data gain calculation and hereditary calculation .The Multi Layer Perceptron (MLP) managed NN is utilized to prepare the significant highlights alone in our proposed framework. This framework improves the Detection Rate (DTR) for a wide range of assaults when contrasted with Intrusion identification framework which utilizes all highlights and chose highlights utilizing hereditary calculation with MLP NN as the classifier. Our proposed framework results, in distinguishing interruptions with higher exactness, particularly for Remote to Local (R2L), User to Root (U2R) and Denial of Service (DoS) assaults.

Published
2020-04-04
How to Cite
Mohith Sai Krishna.Katakam , Komali.Devineni, Pavana.Kanagala, DR.N.Raghavendra sai. (2020). ANALYSIS OF ARTIFICIAL NEURAL NETWORKS BASED INTRUSION DETECTION SYSTEM. International Journal of Advanced Science and Technology, 29(5s), 928 - 935. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/7832