A Hybrid Model for Parameter Tuning Using SSO and CNN

Social Spider Optimization (SSO) and Particle Swarm Optimization (PSO) algorithms applied to Intrusion Detection Systems (IDS) across small and large datasets. The study aims to evaluate and compare both PSO and SSO and find best optimal parameters in order to improve the accuracy of IDS. Additionally, the SSO algorithm find the best hyper-parameter settings which are kernel, filter, padding, learning rate, and stride to achieve optimal results that directly impact IDS security. The evaluation of algorithms in terms of accuracy, precision, recall, F1 score, and AUC score. Results indicate that, on small datasets, PSO comes under overfitting but not SSO. Even on larger datasets, SSO crosses PSO significantly in terms of accuracy. This finding shows the importance of improvement in scalability, efficiency, and robustness of IDS with help of Social Spider Optimization (SSO).