Evaluating Sequential Combination of Two Genetic Algorithm-Based Solutions for Intrusion Detection

摘要: The paper presents a serial combination of two genetic algorithm-based intrusion detection systems. Feature extraction techniques are deployed in order to reduce the amount data that system needs process. designed is simple enough not introduce significant computational overhead, but at same time accurate, adaptive and fast. There large number existing solutions based on machine learning techniques, most them high overhead. Moreover, due its inherent parallelism, our solution offers possibility implementation using reconfigurable hardware with cost much lower than one traditional model verified KDD99 benchmark dataset, generating competitive state-of-the-art.