THREAT INTELLIGENCE FOR CYBER ATTACK PREDICTION

Abstract

This respective research examined threat intelligence (TI) as the proactive ability to identify and respond to cyber threats and highlights machine learning as the means to bolster cybersecurity. As digital infrastructures increase so does the potential for risks such as ransomware DDoS APTs etc. In most cases, established approaches to cybersecurity fail at detecting such threats; thus, turning to a predictive model. In this paper, different machine learning algorithms such as neural networks, decision trees, random forest, and Support Vector Machine are discussed to enhance the threat detection process. Thus, Neural networks as far as accuracy are the best when it comes to identifying such special features which are crucial in identifying new kinds of threats. On the other hand, random forests offer a nice middle ground in terms of accuracy ensuring that there is explainable adherence to compliance requirements such as in the financial or health care sectors. The research also explores areas of specific interest regarding the applicability of TI within different sectors, and its effectiveness. The results emphasize one method for future enhancements in TI: improved data-sharing procedures, ethical artificial intelligence, and models that can endure novel threats. To the best of the author’s knowledge, this study fills this gap through advancing the framework for a multi-layered TI approach and improving the advancement of organizations’ capabilities to appropriately address cyber threats within a constantly evolving digital terrain.