This paper explores the application of Symbolic Regression (SR) to data derived from the Okumura propagation model for predicting path loss in wireless systems. Traditional machine learning techniques, such as artificial neural networks and support vector regression, lack the ability to provide mathematical expressions, limiting physical insights. In contrast, SR can generate analytical expressions, offering a more interpretable model. The study compares the SR-derived expression with the Hata model, which also originates from Okumura. Results indicate that the SR model achieves a better fit to the Okumura data, with a RMS error at least 2 dB smaller than that of the Hata model. This demonstrates the potential of SR in enhancing the accuracy and interpretability of path loss predictions.