Prediction of the parking lot user type : Applying AutoML

Abstract

The process of building a high-quality predictive model for a specific task involves data preprocessing, feature selection, model selection, model hyperparameter optimization, critical analysis of results, and model deployment. These require a lot of human effort and expertise. To alleviate this problem, there is a need for automated machine learning (AutoML) that can be easily performed without the need for specialized knowledge. This study compares the prediction accuracy of the automated machine learning model with the conventional prediction model by estimating the parking lot user type. In this experiment, the AutoML model achieved 0.6543 accuracy.