Machine Learning Models
Intrusion Detection Systems
As network technology rapidly increases, the security of this technology has become a survival need for an organization. Computer and internet infrastructure systems are widely used in many security–critical systems such as banking, finance, education, health, electronic government applications. Due to the rapid growth in technology and widespread use of the internet, many problems are encountered for the system to secure critical information within or between networks. Millions of people attack systems to obtain essential information or damage them. Intrusion detection and prevention systems play a crucial role against these attacks by protecting the critical knowledge of the system. Since firewalls and anti–viruses are not enough to provide complete protection to the system, organizations must implement intrusion detection and prevention systems to protect their critical information against various types of attacks.
Intrusion Detection Systems are one of the most critical security components, covering methods that monitor the activities on the network and analyze the traffic to detect possible attacks, violations, and threats. On the other hand, intrusion prevention systems are network security systems that cover the detection and prevention of seizures.
Today, at the same time, these complex network systems can no longer be protected only with encryption or firewall; realities such as network traffic are constantly mat monitored. Hence, made accurate–time detection of attack attempts is inevitable.
These systems can also reveal weaknesses in the security policies of companies and institutions. Furthermore, intrusion prevention systems also detect the information gathering activities of the attackers about the network and prevent the attackers from getting information about the plans. Therefore, they can also stop it at this stage.
Intrusion Detection and Prevention Systems should be strategically positioned within the network against network attacks. Thus, any attack or abnormal network traffic can be detected quickly. The detection and response teams can take the necessary measures to cyber incidents before the incidents grow.
Types of Intrusion Prevention System
1) Intrusion Detection System (IDS)
It creates and sends alarms on itself, detects violations, and blocks source traffic.
2) Intrusion Prevention System (IPS)
This system enables the production of more detailed alarms and records through a cyber incident management system installed on the network, enabling the detection and response teams to monitor and attack in more detail and take precautions against cyber threats.
Common Detection and Prevention Techniques
1) Signature Based Technique
This system proceeds through comparing signatures with observed events to identify possible events. An alarm is generated if there is a successful match.
2) Anomaly Based Technique
An intrusion detection and prevention system using anomaly–based techniques have profiles that represent the expected behavior of users, hosts, network, and applications.
3) Stateful Protocol Analysis Technique
Stateful protocol analysis compares generally accepted definitions of correct protocol activity with observed events for each protocol state.
4) Intrusion Detection System with Machine Learning
In Machine Learning, classification algorithms learn the output from the given training set inputs and then try to correctly classify the test data for which the class is uncertain.
The KDD–99 dataset is widely used for intrusion detection. This dataset includes a wide variety of attacks simulated on a military network.
Click for more information on this topic;
