Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/250520
Title: | Effective Use of Data Mining in Intrusion Detection |
Researcher: | Vinila Jinny S |
Guide(s): | Jayakumari J |
Keywords: | Engineering and Technology,Computer Science,Computer Science Information Systems |
University: | Noorul Islam Centre for Higher Education |
Completed Date: | 04/07/2016 |
Abstract: | ABSTRACT newlineComputer systems connected through network plays an important role in day to day life. Data transmission through network has increased from kilo bytes to terra bytes. People in large enterprises transmit data for their business dealings, People at academic side transmit data to share knowledge, experts at various departments share their conclusions etc. All these will be targeted by the hackers, who disturb the normal transmission of data. Hence an effective mechanism to identify such a misbehaviour activity is needed. There are intrusion prevention systems and various authentication methods to tackle this problem. All these mechanisms face various drawbacks in identifying the hackers. With these challenges, Intrusion detection system (IDS) evolved, where the system analyzes the audit data for intrusion. The performance of the IDS can be improved by better analysis of the audit data. The analysis task can be better handled by data mining algorithms. The main aim of IDS is to have reliable detection and cope with large amount of network traffic. newlineThis research aims to develop a novel hybrid systematic framework to semi-automate the process of building intrusion detection systems. A basic premise is that when audit mechanisms are enabled to record system events, distinct evidence of legitimate and intrusive activities will be manifested in the audit data. In this research work intrusion detection is considered as a data analysis task. The objective is to design a high performance IDS with the help of data mining algorithms, which improves the accuracy and efficiency of the overall IDS and reduces False positive rate and False negative rate. newlineIDS effectiveness can be improved by developing an accurate detector module. The accuracy of the detector module is decided by the detecting patterns stored in the knowledge base. Detecting patterns can be generated either by supervised learning and unsupervised learning method. Unsupervised Learning methods learn with instant pattern but it may result in high False Nega |
Pagination: | 131 |
URI: | http://hdl.handle.net/10603/250520 |
Appears in Departments: | Department of Computer Science and Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
acknowledgement.pdf | Attached File | 7.47 kB | Adobe PDF | View/Open |
certificate.pdf | 15.41 kB | Adobe PDF | View/Open | |
chapter iii.pdf | 372.73 kB | Adobe PDF | View/Open | |
chapter ii.pdf | 324.47 kB | Adobe PDF | View/Open | |
chapter i.pdf | 35.83 kB | Adobe PDF | View/Open | |
chapter iv.pdf | 357.31 kB | Adobe PDF | View/Open | |
chapter vii.pdf | 49.66 kB | Adobe PDF | View/Open | |
chapter vi.pdf | 706.07 kB | Adobe PDF | View/Open | |
chapter v.pdf | 308.25 kB | Adobe PDF | View/Open | |
references.pdf | 233.22 kB | Adobe PDF | View/Open | |
title page.pdf | 17.95 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: