Mobility on Demand transforms the way people travel in the city and facilitates real-time vehicle hiring services. Given the predicted future travel demand, service providers can coordinate their available vehicles such that they are pre-allocated to the customers' origins of service in advance to reduce waiting time. Traditional approaches on future travel demand prediction rely on statistical or machine learning methods. Advancement in sensor technology generates huge amount of data, which enables the data-driven intelligent transportation system. In this paper, inspired by deep learning techniques for image and video processing, we propose a new deep learning model, called Multi-Scale Convolutional Long Short-Term Memory (MultiConvLSTM), by considering travel demand as image pixel values. MultiConvLSTM considers both temporal and spatial correlations to predict the future travel demand. Experiments on real-world New York taxi data with around 400 million records are performed. We show that MultiConvLSTM outperforms the existing prediction methods for travel demand prediction and achieves the highest accuracy among all in both one-step and multiple-step predictions.
Travel Demand Prediction using Deep Multi-Scale Convolutional LSTM Network
2018-11-01
496020 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Travel time prediction with LSTM neural network
IEEE | 2016
|Network Scale Travel Time Prediction using Deep Learning
Transportation Research Record | 2018
|