This study examines how improving the predic-
tion of cyberattacks can be achieved by combining
predictive modelling with Network Intrusion Detection
Systems (NIDS). Proactive detection is essential for efficient
cyber security as cyber threats change. We offer a system
for analysing real-time network behaviour from NIDS and
historical attack data using machine learning. Our method
increases accuracy and reaction times by emphasising
feature selection, data preprocessing, and different
predictive models. According to experimental findings, this
in- tegrated approach performs noticeably better than
conventional detection methods and offers early warnings
of possible hazards. With the help of this framework,
organisations can improve situational awareness and lessen
the effects of cyberattacks.