I am working on NSL-KDD data set for my research work. I want to separate all 41 attributes except 42nd class attribute into four class labels like Basic, Content, Traffic and Host. It is not possible in WEKA tool, can anyone help me which tool is available and how.
Hi all Thank you for your response . I have that paper but how to get it is not clear.
Hi,
I cannot understand what you mean. The proper category of the all features were determined by producers. For more detail refer to the Table 1 of the following paper:
Article Hybrid of binary gravitational search algorithm and mutual i...
Hi,
you can use a meta heuristic algorithm :Biogeography-based optimization (BBO), Genetic algorithm,Bee swarm optimisation Algorithm (BSO). You can implement those algorithms using Weka is a good tools i used it on NSL KDD.
please read this abstract :
With the increasing deployment of computer networks and the intensive use of the Internet, computer security is not only a need but also a necessity for organizations that offer their services via the internet and are constantly targets of malicious activity. An intrusion detection system (IDS-Intrusion Detection System) aims to identify and respond to malicious activities targeting a computer or computer network. The problem of intrusion detection returns to classify the flow of system activity into two categories: normal activity flow or stream of attacks. The problem thus formulated can be considered as a classification problem, which can itself be formulated as an optimization problem in which the objective function is to maximize the detection rate and the solution is a partition. Furthermore, the selection of attributes, which is a task of data mining to reduce the dimensionality in the data set by selecting the most relevant attributes, may been formulated as a combinatorial optimization problem. Indeed, applied to intrusion detection, this problem would be the objective function to maximize detection rates and as space solutions in all possible sets of the set of attributes of departure. Therefore, metaheuristics can be used to solve this problem. The work required under this project graduation is to apply metaheuristics BSO (Bee Swarm Optimization), BBO (Biogeography-Based Optimization) and Genetic Algorithm GA to the problem of feature selection for the optimal subset of attributes that will be considered in the classification stage. Experiments were conducted throughout KDD'99 data and comparison between metaheuristics were done. We have also make comparison between wrapper approach and filter approach.
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining