An efficient cyber threat prediction using a novel artificial intelligence technique

An efficient cyber threat prediction using a novel artificial intelligence technique

연구 분야: Safety

학회: Multimedia Tools and Applications

Digital applications are ruling today's world with their advancement. However, offering security for that digital application is an important and complex task. Several detection-based security models have existed in the Artificial Intelligence (AI) vision. Still, the problem in threat detection has not ended because of the unique behavior of the different attacks. So, the present research has introduced a novel Cuttlefish-based Peephole Long Short Term Memory (CbP-LSTM) model proposed for predicting the cyber threat from the data defends against attacks. Initially, data were preprocessed by removing noise from the data using the noise filtering function. Then, the refined data is imported to the classification layer of the CbP-LSTM for performing the feature extraction and attack prediction tasks. Moreover, the proposed CbP-LSTM model was implemented in the Python tool with several performance metrics, whereas the parameters were calculated, such as accuracy, precision, Recall, and F-score. This proposed model produced outstanding results compared with the previous work by providing the highest predicted accuracy of cyber threats.