Intrusion Detection using Naïve Bayes

In This project there are 2 types of input we are giving. One is online in which we are capturing the packets online using packet sniffer tool (Packets are of any kind of extension). Other one is standard NSL-KDD dataset (testing dataset).

After capturing the Packets (inputs) then we are performing pre-processing. In pre-processing we are removing

1) blank spaces( 0,0,0.00,0,1,__,0,0.00) here red lines denotes blank space

2) If in the one record it contain symbol, Character like @,!,#,$...etc then we have to remove it.

3) The duplication of record like

Ex. tcp,1,0,0,http,1,0,0,0,3,2,1

tcp,1,0,0,http,1,0,0,0,3,2,1

We have to take the similar record only once.

After this we got pre-processed data. This pre-processed data should be in MY-SQL Database Table.

On the pre-processed data we are going to apply 3 FST techniques. Then we got output as a set of reduced features.

Simultaneously on the same pre-processed data (Not on the output of FST techniques) we are going to apply FVBRM. Before going for Naive Bayes classifier we have to calculate classifier’s accuracy, true positive rate, false positive rate, root mean square error. Then Naive Bayes classifier is applied and then we got output as set of reduced features separately.

In the last module that is evaluation we are going to compare the output of FST and FVBRM. And then we got efficient reduced set of features that will be the output of our project which will be used for implementation of Intrusion detection system.

This above Project Should developed in JAVA only without using external API or Tools like weka etc.

Java MySQL

Numer ID Projektu: #7097951

O projekcie