.png){kind=icon}
Return
Goat house swallow algorithm based deep learning for intrusion detection in cloud computing with MapReduce framework
.png){kind=icon}
PRE
AI
DOI：10.1016/j.compeleceng.2026.111476.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Intrusion detection system (IDS) is an important security solution in a cloud computing environment, enabling earlier detection of attacks and malicious actions. By utilizing the data offered by IDS, it is probable to apply suitable measures and mitigate attacks that would otherwise severely weaken network security. The conventional IDS confront limitations in assessing big data for identifying an anomalous behavior intelligently. In this research, the Goat House Swallow Algorithm based Deep Convolutional Neuro-Fuzzy Inference System (GHSA_DCNFIS) is designed for intrusion detection in cloud computing. At first, the cloud system model is simulated, after which the input log file is obtained from the datasets. Intrusion detection is then performed using the MapReduce framework, which includes both mapper and reducer phases. The data is normalized using the hyperbolic tangent (tanh) function, and relevant features are chosen in the mapper stage through deep tabular learning (TabNet). During the reducer stage, data is augmented based on the K-means-Synthetic Minority Over-sampling Technique (SMOTE). Lastly, intrusion is detected utilizing Deep Convolutional Neuro-Fuzzy Inference System (DCNFIS). Moreover, DCNFIS is trained employing the Goat House Swallow Algorithm (GHSA). However, GHSA is introduced by integrating the Goat Optimization Algorithm (GOA) with the House Swallow Optimizer (HSO). Additionally, GHSA_DCNFIS has acquired an accuracy of 96.743%, a detection rate of 98.399%, an F1-score of 95.292%, a precision of 95.246% and a recall of 95.338%.
Keywords:
Cloud computing
Intrusion detection
MapReduce framework
Deep convolutional neuro-fuzzy inference system
Goat optimization algorithm
Journal
C
IF:
4.9
Papers:
248
Citations:
0
Organization
Cited Papers
No cited papers available