Due to the devastating effects of various types of attacks, the protection of the Industrial Internet of Things (IIoT) is crucial. Machine Learning is an effective tool for analyzing and safeguarding IIoT. In this work, the authors used various Machine Learning models to distinguish normal traffic from cyber-attacks and compared their performance. The authors proposed a hybrid model with XGBoost Classifier and Random Forest Classifier which is observed as the most effective with an accuracy of 99.996%. The performance of the proposed model is evaluated using representative metrics for better model evaluation.