In the constantly changing field of cybersecurity, advanced and proactive defense mechanisms are crucial. This study investigates the utilization of Artificial Intelligence (AI) and Machine Learning (ML) in predictive analytics to forecast and categorize various cyberattacks, with a specific focus on Distributed Denial of Service (DDoS) and Man-in-the-Middle attacks. The study utilizes a detailed Edge-Industrial Internet of Things (IIoT) dataset that represents actual network traffic situations. The proposed method combines Isolation Forest, Random Forest, and One-Class Support Vector Machine (SVM) to improve the accuracy of attack prediction using an ensemble approach. Isolation Forest is effective at identifying anomalies in datasets, Random Forest is strong at classification, and One-Class SVM is adept at identifying normal behavior patterns. These techniques work together to create a comprehensive and efficient defense system against various cyber threats. The predictive analytics model evaluation shows an accuracy rate of 98.23%, highlighting the system’s capability to classify incoming network traffic as normal or indicate an attack with precision. The True Positive Rate (TPR) reaches a notable 98.78%, demonstrating the model’s accuracy in detecting and predicting cyberattacks. The proposed ensemble method demonstrates a high level of accuracy and sensitivity, highlighting its potential practicality and reliability in real-world cybersecurity applications. The study enhances the development of sophisticated predictive analytics models for cybersecurity and emphasizes the importance of using AI and ML techniques, along with diverse datasets, to strengthen and adapt cyber defense mechanisms against changing cyber threats.