Enhance of Extreme Learning Machine-Genetic Algorithm Hybrid Based on Intrusion Detection System

摘要

This study presents a new scheme of the hybrid Extreme Learning Machine-Genetic Algorithm (ELM-GA). ELM has been proved to be exceptionally fast and achieves more generalized performance for learning Single hidden Layer Feedforward Neural networks (SLFN). However, due to the random determination of parameters for hidden nodes and the number of hidden neurons, some un-optimal parameters may be generated to influence the generalization performance and stability. Some of the papers used GA as a hybrid to solve this problem in ELM but ELM-GA still has some limitations where they used the GA to find the optimal weights for the ELM. In this research, we try to let the GA not only find the best weights but find the best classifier (weights and structure). Intrusion Detection System (IDS) facing big challenge in high rate of false alarms. This research proposes a new method in validation of the classifiers to be sure that the classifiers training enough to mitigate the false alarm?s rates.