Cyber Threat Detection Based on Artificial Neural Networks
P Ramya Sai, K. S. Niraja · International Journal for Research in Applied Science and Engineering Technology · 2023
Abstract: Finding an automated method for detecting cyber-attacks is one of the biggest problems in cybersecurity. We describe an artificial intelligence (AI) method based on deep learning for detecting cyber threats. The suggested solution uses a deep attempting to learn monitoring method to improve cyber-threat identification by breaking down a large volume of recorded security events into event profiles. We created an AI-SIEM system that combines event profiles for data pretreatment with several multilayer perceptron techniques, such as FCNN, CNN, and LSTM. The approach focuses on separating real positive warnings from misdiagnosis alerts to assist experts in quickly responding to cyber-attacks. We carried out tests utilising the five traditional machine-learning techniques to assess the comparison study with existing approaches (SVM, k-NN, RF, NB, and DT). The experimental findings of this study confirm that our suggested methods may be used as studying models for activity recognition and demonstrate that, when used in the real life, they outperform traditional auto techniques.