IDO-DKNN: NETWORK ATTACK DETECTION AND MITIGATION VIA OPTIMIZED DEEP KRONECKER NEURAL NETWORK

Complex attack vectors and ever-evolving threats continue to bother network security. Accordingly, effective detection and mitigation approaches have been essential to protect modern digital infrastructure. In turn, this work tries to solve the challenges by means of a comprehensive methodology based on integrating advanced techniques in data preprocessing and feature extraction with an optimized deep learning model. This involves strict preprocessing of the data, which includes cleaning to handle redundancies and missing values, standardization to scale the data uniformly, and the extraction of features based on measures of central tendency, dispersion analyses, and information-based metrics. The significant features selected with Enriched Coati Osprey Algorithm (ECOA) were employed for training Optimized Deep Kronecker Neural Network (Opt-DKNN) -so as to identify the attack's effective detection. This approach improves network security through detecting and mitigating attacks with more appropriate use of advanced feature extraction and optimization techniques. The weighted BIAT model is presented for the mitigation of the attack.