| Abstrakt: |
An Innovative intrusion detection system is well defined for the network security which monitors the attacks in the network. The results of the experiment proves that the proposed model has higher accuracy than the existing model. In this study, two groups are used such as the K-Neighbor algorithm and Logistic Regression algorithm. Dataset is collected which consists of 22,545 records. The algorithms were developed, tested, and trained on the dataset. In this study, the sample size will be iterated twice. Iteration 1 uses the Train Set, while Iteration 2 uses the Test Set. It takes ten cycles to figure out which data is anomalous and which is typical. When doing iterations and implementi ng algorithms, G-Power is utilised to the tune of 80%. The accuracy of the Logistic Regression algorithm is 95.45, while the K - Neighbor algorithm is 99.87, according to the findings of the trial. Between the accuracy of two algorithms, there is a considerable statistical difference. Using independent samples t-tests, the statistical difference is p0.05. With the illustration of algorithms applied to the data set in testing its hypothesis, this study explains the process for using logis tic techniques and KNN. The major goal is to create a cutting-edge intrusion detection system that can classify whether data is anomalous or not. When compared to the Logistic Regression Algorithm, the results suggest that the K -Neighbor Algorithm performs better. [ABSTRACT FROM AUTHOR] |
