We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Fuzzy K-Means with M-KMP: a security framework in pyspark environment for intrusion detection.
- Authors
Begum, Gousiya; Ul Huq, S. Zahoor; Kumar, A. P. Siva
- Abstract
In recent times, IDS (Intrusion Detection System) has become a significant tool for improvising network security through the detection of abnormal and normal data. It is vital as it permits one to identify and respond to incoming malicious traffic. The intruders have also enhanced the inclusion of attacks in systems with a recent increase in data. Concurrently, ML (Machine Learning) algorithms can learn from corresponding data that has been afforded. With the provision of new data, the accuracy and efficacy of the ML model to take decisions to enhance with training. However, with the evolution of big data, ML has turned incapable of handling huge data interpretation issues which made most of the conventional systems explore high FP (False Positive) rates and low accuracy rates. This gave rise to pyspark which serves as a platform for addressing these issues that the ML method fails to solve. ML in pyspark is a scale and easy to use. Considering this, the present research intends to propose ML-based algorithms for classifying intrusion detection in a pyspark environment. This study proposes a security framework named Fuzzy K-Means with M-KMP (Modified-Knuth Morris Pratt) wherein the clustering is accomplished by Fuzzy K-means which is capable of exploring data points that potentially relate to multiple clusters. Whereas, M-KMP achieves information matching on the clustered data for assessment of the information occurrence on the allocated threat data that will serve as an assistance for security developers in attack prevention. The efficiency of this proposed work is confirmed through the results.
- Subjects
MACHINE learning; COMPUTER network security; ALGORITHMS; INTRUSION detection systems (Computer security); INSTRUCTIONAL systems
- Publication
Multimedia Tools & Applications, 2024, Vol 83, Issue 30, p73841
- ISSN
1380-7501
- Publication type
Article
- DOI
10.1007/s11042-024-18180-5