Machine Learning Project for Network Intrusion Detection

Project Discription

Clustering of nodes were done by Wolf optimization technique, algorithm is able to provide solution dynamic situation. Cluster nodes selection done on the basis of device energy, distance from the base station. The fitness of the cluster centre was determined by taking into account the distance between nodes and their energy levels. Routing of packet is also done from the nodes by means of cluster centers. In order to reduce the load of cluster nodes, shuffling of nodes were done time to time. Experiment was done on different environment of underwater and varying number of nodes. Network increases the use of electronic devices and provide support for various industries. Many of intruders are taking advantages of weak security. This paper has proposed a model that detects the malicious session in the network and generate alarm for that. Proposed model uses artificial immune system genetic algorithm  for feature reduction and train the neural network for detection. Experiment was done on real dataset obtained from [14]. Result shows that precious value was enhanced by 12.1% and accuracy of the model was enhanced by 8.86% as compared to existing models. In future scholar can train other machine learning models for getting better results.

Clustering of nodes were done by Wolf optimization technique, algorithm is able to provide solution dynamic situation. Cluster nodes selection done on the basis of device energy, distance from the base station. The fitness of the cluster centre was determined by taking into account the distance between nodes and their energy levels. Routing of packet is also done from the nodes by means of cluster centers. In order to reduce the load of cluster nodes, shuffling of nodes were done time to time. Experiment was done on different environment of underwater and varying number of nodes.

Project Sample Image