This bachelor’s thesis focuses on the application of machine learning methods in the field of Operational Modal Analysis (OMA). The introductory part provides a review of traditional approaches to modal analysis, namely numerical modeling using the Finite Element Method (FEM), Experimental Modal Analysis (EMA), and Operational Modal Analysis (OMA), including their main advantages and limitations from the perspective of engineering practice. Subsequently, modern data-driven methods, particularly Koopman theory and the Dynamic Mode Decomposition (DMD) method, are introduced, and their connection with machine learning algorithms is discussed. The main contribution of the thesis is the design and implementation of a PhysicsInformed Neural Network (PINN) model intended for the identification of the dynamic properties of a cantilever beam. The proposed model combines the ability of neural networks to approximate measured data with physical constraints derived from the Euler–Bernoulli beam vibration theory. The thesis also includes the design of the neural network architecture, the formulation of a physics-informed loss function, and the implementation of advanced training strategies such as Curriculum Learning and adaptive balancing of individual loss-function components. The functionality and accuracy of the developed model are validated using synthetically generated data, and the results are compared with the outputs of the established PyOMA library. The results demonstrate that incorporating physical laws into the learning process improves the robustness of modal parameter identification and represents a promising approach for the field of Structural Health Monitoring (SHM).