Efficient Intrusion Detection based on Kulczynski Similarity Measurement

In this study, to facilitate the identification of hazardous software and attempts to compromise it, it is recommended that a modified version of the Kukeyzy similarity measure be utilized with the assistance of cybersecurity experts. This suggestion is based on the findings of this study. Block-wise similarity partitioning was created with the primary purpose of achieving a high degree of detection accuracy while also reducing the number of false alarms sent out. The goal of creating this plan was to make this potential a reality. We used both the UNSW-NB15 benchmark dataset and the NSL-KDD dataset. We can access both of these datasets. We demonstrated that our model outperforms traditional methods using well-established criteria, including accuracy, precision, recall, and F1 score. These criteria were used to do this. As a result, we successfully executed what we set out to do. The model performed exceptionally well, particularly in accurately predicting 98.47% of the data associated with the UNSW-NB15 and 93.18% of the data linked to the NSL-KDD. This indicates that the model performed well. Both of these numbers show that the product is of good quality overall. The method is straightforward to understand and works effectively, which makes it a good choice for cybersecurity applications.