Enhancing Intrusion Detection with Grey Wolf Optimization-Tuned Neural Networks in a Big Data Framework

Nowadays, the effective intrusion detection and study of secure landscapes in big data environment faces many challenges. This work intends to employ an intrusion detection scheme, which encompasses three phases “(i) Preprocessing, (ii) Feature Extraction, and (iii) Intrusion Detection”. Initially, the input data is pre-processed with hadoop and map reduced frameworks that aids in the process of data normalization. Then, from the normalized data, flow based features are extracted that are then given to the intrusion detection phase. Optimized Neural Network (NN) is used for this purpose. For more precise prediction, the training (weight) of NN is done by”Grey Wolf Optimization (GWO) Algorithm” via tuning the optimal weight. Finally, the superiority of the developed model is assessed by comparison with existing methods.