Long Range (LoRa) technology has emerged as a promising solution for Low-Power Wide-Area Networks (LPWANs) in applications ranging from smart cities to environmental monitoring. This paper explores the integration of machine learning algorithms to predict LoRa network efficiency in smart traffic applications. By making use of machine learning, we predict and optimize network settings, including Spreading Factor (SF), Coding Rate (CR), and Bandwidth (BW). Our machine learning model, employing the RandomForestRegressor algorithm, accurately predicts network efficiency based on input parameters such as Energy, DER (PDR), Throughput, Bitrate, and Delay. Furthermore, the machine learning model predicts and categorizes network efficiency levels as “High,” “Average,” and “Low” for informed decision-making. Comparative analysis reveals the performance of various machine learning algorithms across different environmental conditions, with Random Forest Regressor and Linear Regression standing out as high-accuracy options. These findings provide valuable insights into the integration of machine learning to enhance LoRa network efficiency, ultimately leading to improved scalability, adaptiveness, and traffic management in smart cities.