Car Parking is a significant issue in urban zones in both developed and developing nations. Following the quick incense of vehicle possession, numerous urban communities are experiencing a lack of car parking regions. Keeping in mind that issue, we undergo the study of Frankfurt Car Parking Data. The aim of this research work is to test several prediction strategies to provide the users with information about occupancy rates of parking lots. With the approach of self-sufficient vehicles as the future and automatic car parking features in cars, realizing the occupancy rates of a parking area beforehand can be valuable and can spare a ton of time and fuel. To predict occupancy rates we are using the following prediction models namely Linear Regression, Neural Networks, Support Vector Regression, Decision Trees (Regression) and Ensemble Decision Trees. We have also implemented K-means clustering as we hypothesise that adding one more feature to the dataset for similar instances would help predictive algorithms to fit better on the data. Using this additional feature, we modified the existing dataset D1 (with 3 features) into D2 (with 4 features). We advocate this hypothesis by comparing the results of prediction algorithms on both datasets (D1 and D2). From the results, we found out XGBoost fits the dataset exceptionally well. © 2020, Springer Nature Singapore Pte Ltd.