IoT-Enabled Smart Parking System using Machine Learning for Real-Time Parking Prediction

Because of the growing number of cars and the inadequate parking infrastructure, urban parking management has become more and more crucial. To predict parking spot availability in IoT-enabled intelligent systems, this study assesses the performance of many machine learning (ML) models, including decision tree (DT), random forest (RF), extreme gradient boosting (XGBoost), and extra trees classifier (ETC). F-score, accuracy, recall, and precision were used to gauge performance. The DT model's 88.63% accuracy and 87.61% F-score demonstrated its low predictive power. With an F-score of 91.45% and greater accuracy of 92.30%, RF proved appropriate for medium-sized facilities. XGBoost performed better, capturing intricate parking patterns with 93.87% accuracy and 93.50% F-score. With 93.00% accuracy and 92.74% F-score, ETC performed similarly to XGBoost and provided computational advantages for big datasets. ETC and RF are dependable in medium-sized settings, however XGBoost is advised for large-scale settings.