Application of revised firefly algorithm and grey wolf optimisation on keystroke dynamics

In this digitalised world, to countermeasure computational threats, keystroke dynamics (KD) is one potential biometric feature that is used to enforce security over a network. Feature subset selection (FSS) process further aid for the increase of security by selecting the appropriate features, which make the replication of pattern difficult. For this purpose, two commonly known algorithms namely firefly algorithm (FA) and grey wolf algorithm (GWA) are being enhanced by incorporating chaos and pheromone in the network architecture. The experimental results have shown the robustness of the revised algorithms of firefly and grey wolf where optimum values for false acceptance rate (FAR) are being achieved. Besides, this study has shown that the revised FA fits better as a FSS technique by outperforming previous proposed solutions in terms of recognition rate (RR), where above 95% has been achieved, which shall aid in the reduction of attacks.