원문정보
초록
영어
An Intrusion Detection System is an application which observes movements or action happen on the network and determine it for any kind of harmful activity that can disturb computer security policy. With progress of increase the usage rate of the internet, there is a widely increase in the number of internet attacks as well, so contests arise towards the network security due to the arrival of new approaches of attacks. To classify these attacks, a new hybrid method with the help of data mining based on decision tree C4.5 and Meta algorithm is planned. This method gives a classifier which expands the whole accuracy of detection. Many data mining techniques have been settled for detecting intrusion. For recognition of anomalies a hybrid technique based on decision tree C4.5 with Meta algorithm is offered that provides better accuracy and reduces the problem of high false alarm ratio. The assessment of the given approach is made with other data mining techniques. With this given approach detection rate is improved significantly. KDD Cup 1999 dataset use for experimental work.
목차
1. Introduction
2. Literature Survey
3. Intrusion Detection System
3.1 Type of Alerts
3.2. Type of IDS
3.3. Types of Detection
4. Anomaly/Statistical based Detection
4.1. Supervised Learning
4.2. Unsupervised Learning
5. Objectives
6. KDD Cup 1999 Dataset
7. Pre-processor
8. Proposed Algorithm
9. Experimental Results
9. 1. Accuracy
9.2. Detection Ratio
11.3. True Positive Ratio
9.4. False Positive Ratio
9.5. Precision Ratio
9.6. Recall
9.7. F-Measure
9.8. Area under the ROC Curve
9.9 Area Under the PRC
10. Conclusion
11. Future Scope
References