Analysis Of Intrusion Detection System With Machine Learning And Deep Learning Techniques

To secure a network from Intrusions,an Intrusion detection system plays a significant role in maintaining Confidentiality of any data in any Organization.We achieved our main goal for this paper through confusion matrix and share the results through precision recall and F1 score.In this paper,we have discussed and explore the accurate performance of a NIDS which can detect many types of attacks in the network using Neural Network Algorithm.We have analyzed the accuracy of our model through comparison with different types of attacks.Our main objective is to detect different types of network attacks without contingent upon the past experience and its first attempt.We have used CICIDS2017 data set in this paper to evaluate the proposed model,we used all illustration in the dataset for comprehensive detection.We used dataset for reducing the false alarm rate.The Neural network improves the accuracy and performance for Intrusion Detection System and provides a new means as a research method for intrusion detection in any Network security appliance.We discuss different algorithms of deep learning and machine learning in this paper.We will address the challenges for detecting intrusion method,and will provide suggestions for these methods.In this paper we just used one technique that is Neural Network for binary and multi-classification problem for the dataset of CICIDS2017.There is lot of steps that we prepare the data for our deep learning model.First,we clean it through some machine learning techniques and we selected most important features from the dataset and then we applied min max normalization technique on the data.