Prediction Technology for Parking Occupancy Based on Multi-dimensional Spatial-Temporal Causality and ANN Algorithm

摘要

The Granger causality model is extended by supplementing spatial, weather, and other factors. Therefore, a multi-dimensional spatial-temporal causality model for the prediction of the parking occupancy is proposed, and the prediction algorithm for parking occupancy based on multi-dimensional spatial-temporal causality and ANN is carefully designed. The CityPulse dataset provided by the European Union FP7 project is introduced to train the network, and verify our algorithm. The experimental results show that our new technology for prediction of parking occupancy can effectively improve the accuracy of the prediction, compared with other algorithms only rely on time or spatial factors.