Protection against Hardware Trojans (HTs) is gaining importance as integrated circuits (ICs) continue to expand. This thesis represents a hybrid framework of detection that is a fusion of traditional HT-detection methods and machine learning (ML)-based methods to enhance the results of detection in terms of accuracy, speed, and resilience. A large number of classifiers Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Naive Bayes (NB), Gradient Boosting (GB), AdaBoost, Multi-Layer Perceptron (MLP), XGBoost and a Meta-Learning ensemble were compared on a realistic Trojan dataset. A variety of tree-based and ensemble models (DT, RF, GB, AdaBoost) and the Meta-Learning ensemble scored almost perfect on the tests. The results showed that a combination of these ML methods and pre and post fabrication standard checks provided a scalable and workable pipeline of detecting IC security. It was found that the hybrid methodology significantly increased the detection rates compared to the standalone methodologies. In addition, a high-performance direction was achieved with respect to the implementation in production test settings.