Electrical power systems around the world are currently operating near the limits of security and stability design, due to economic considerations and high rates of demand for electrical energy. During their operation, these systems are exposed to various types of disturbances, which in severe cases lead to their complete loss and complete shutdown. Dynamic security assessment is one of the most important tools used in evaluating the performance of electrical power systems after these disturbances occur. The traditional mathematical methods for evaluating dynamic security require complex computations and computational time so that they are not suitable for evaluating dynamic security in real time. To address these challenges, this research was conducted to provide new tools based on advanced techniques of artificial intelligence techniques capable of building a classifier to evaluate its dynamic behavior in real time. The research methodology in this study relied on the use of artificial intelligence techniques, including back propagation artificial neural network, decision tree algorithms (J48) and logistic model tree (LMT). Which was applied to the emergency database of the electrical power system (IEEE 14 Bus) test model after applying the most common types of electrical disturbances. The results of the artificial neural network technology showed high classification accuracy (98.958%) and a lower error rate compared to the decision tree (J48) and (LMT) algorithms. The results of this research are very important to improve the accuracy of the results of the dynamic security assessment classifier for the electrical power system, which makes it easier for the network operator to take appropriate protective measures in the moment of disturbances in the network.