Evaluating APT Detection Through Finance AI and Machine Learning: Is Superior Accuracy Attainable?

EasyChair Preprint 14630

Abstract

Advanced Persistent Threats (APTs) represent a significant challenge in the cybersecurity landscape due to their sophisticated nature and ability to evade traditional detection mechanisms. With the rise of machine learning (ML) as a powerful tool in cybersecurity, this article evaluates the effectiveness of ML-based techniques in detecting APTs and examines whether superior accuracy is attainable. We explore various ML models, their strengths and limitations in APT detection, and the role of data quality and feature selection in enhancing detection capabilities. By conducting a comparative analysis of existing approaches, we aim to provide insights into the potential of ML to improve APT detection accuracy. The findings suggest that while ML offers promising enhancements, achieving consistently superior accuracy requires careful consideration of model selection, training data, and the evolving nature of APTs. The implications for cybersecurity practices are significant, as organizations seek more robust and reliable methods to defend against these stealthy threats.

Keyphrases: APT detection, Advanced Persistent Threats (APTs), Cybersecurity, Model Interpretability, Reinforcement Learning, anomaly detection, feature engineering, machine learning, supervised learning, unsupervised learning

@booklet{EasyChair:14630,
  author    = {John Owen},
  title     = {Evaluating APT Detection Through Finance AI and Machine Learning: Is Superior Accuracy Attainable?},
  howpublished = {EasyChair Preprint 14630},
  year      = {EasyChair, 2024}}