Neural network-based parking system object detection and predictive modeling

<span lang="EN-US">A neural network-based parking system with real-time license plate detection and vacant space using hyper parameter optimization is presented. When number of epochs increased from 30, 50 to 80 learning rate tuned 0.001, the validation loss improved 0.017 training object 0.040. The model mean average precision mAP_0.5 0.988 99%. proposed also uses a regularization technique for effective predictive modeling. modified lasso ridge elastic (LRE) provides 5.21 root square error (RMSE) an R-square 0.71 4.22 absolute (MAE) indicative higher accuracy performance compared other regression models. advantage LRE that it enables via penalty feature selection characteristics both ridge.</span>