Intrusion Detection Data Mining Thesis

Intrusion Detection Data Mining Thesis-4
However, if the attack pattern is slightly altered, this method will not be able to identify the changed versions of the attack [7].An anomaly detection system has a profile of normal behaviour patterns about the defence system.

However, if the attack pattern is slightly altered, this method will not be able to identify the changed versions of the attack [7].An anomaly detection system has a profile of normal behaviour patterns about the defence system.

Tags: References Term PaperInformal Business PlanEssay On Robbery In BankAirport Self Assigned Ip AddressReview Of Related Literature And Studies Research PaperResearch Paper On Gifted StudentsMla Title Of Book In Essay

Anomaly detection techniques demonstrate good accuracy in detecting network-level attacks such as the SYN flood, teardrop, and denial of service (DOS), etc.; but not in recognizing application level exploits such as Remote to Local (R2L) and User-to-Root (U2R).

All of anomaly detection schemes consider only the packet header fields such as flags, port numbers, and IP addresses, etc.; therefore, they work well if an attack involves only the related fields at the network level.

The hybrid intrusion detection method has been developed to improve the performance and capabilities of intrusion detection and prevention systems (IDPS) by combining the signature-based method (misuse detection) and the anomaly-based method [10].

Some evaluation criteria that can be used to compare the performance of algorithms in an IDS include [11]: 1) accuracy, 2) false negative rate (FNR), 3) false positive rate (FPR), 4) time used, 5) memory consumption, and 6) Kappa Statistic.

An intrusion prevention system (IPS) is an advance combination of anti-virus software, personal firewalls, and IDS, etc.

The objective of an IPS is not only to detect attacks, but also to stop them by responding automatically such as disabling connections, logging users offline, ending processes, and shutting the system down, etc.

Current challenges of these methods in intrusion detection are also introduced. In the area of cyber security, effective and efficient situational awareness often requires knowledge of current and historical cyber (i.e.

Cite this paper: Lidong Wang , Randy Jones , Big Data Analytics for Network Intrusion Detection: A Survey, International Journal of Networks and Communications, Vol. host or network) activities to detect and respond to threatening behaviours [2].

Unfortunately, they have no way to detect attacks if the payload is involved, for example, an attack on Microsoft IIS induces users to download a malicious script file, but because there are no invalid packet header fields involved, header-based techniques will not trigger any alarm.

Malicious payload is a type of application level attack; if the payload is ignored in anomaly-based detection, poor performance in detecting payload associated attacks is obvious [9].

SHOW COMMENTS

Comments Intrusion Detection Data Mining Thesis

The Latest from auto-znaniya.ru ©