Intrusion Detection Using Dynamic Feature Selection: An Adaptive Bacterial Foraging Method

The continuous area of cyberattacks demands more responsive defenses than classic, static intrusion detection systems (IDS) can provide. This research provides a unique method that exploits the skills of adaptive bacterial foraging optimization (BFO) for dynamic function selection in IDS. This strategy overcomes the constraints of static strategies by exploiting the collective intelligence and versatility of BFO, permitting real-time choice of the maximum vital attributes primarily based on current feedback and attack patterns. Utilizing chemotaxis, swarming, and elimination/dispersal strategies, the algorithm constantly refines its characteristic choice technique.Extensive critiques on the CICIDS2018 dataset display the prevalence of our adaptive BFO approach. Compared to static BFO and different dynamic techniques, our technique constantly excels across multiple attack types, achieving superior accuracy, precision, and F1 score. Ablation research further emphasizes the relevance of excellent adaptive components, with swarming playing a massive role. By handing over superior detection accuracy, flexibility, and computing economy, the adaptive BFO method enhances IDS.