Optimum Analysis of Imbalanced Network for Intrusion Detection using LSTM Convolution Technique

Abstract

Analyzing network packets to determine whether they are genuine or suspicious are called “Intrusion Detection.” The significant difficulties associated with this space incorporates the tremendous volume of information for preparing and the quick and streaming information that will be accommodated the expectation interaction. In addition, the intrusion detection model faces additional difficulties as a result of the domain's inherent data imbalance. The classification accuracy and other parameters of enhanced LSTM are contrasted with those of conventional deep learning and other machine learning methods in this study. In addition to classifying the tweets, this framework can be used to investigate user attitudes toward Indian higher education. Two algorithms form the basis of the proposed framework: Using the evolutionary algorithm to improve LSTM. Because the standard LSTM algorithm can select parameter values at random, the enhanced LSTM algorithm uses the evolutionary algorithm to enhance its functionality.