predict the spatial and temporal distribution of flight delay
What is it about?
Along with the rapid increasement of flights and projects of extending and building airports, the probability of flight delays is also increasing. People begin to pay more attention to the prediction of flight delays in a large civil aviation air traffic network. In this paper, we employ a deep learning (DL) model-the convolutional long short-term memory network (conyLSTM), to address the airport delay prediction in network structure. The spatiotemporal variables including flight delays of airport, air route congestion, airport throughput and flow control are input into an end-to-end learning architecture as a spatiotemporal sequence. The future flight delays in airport will be output by the model. Experiments show that conv-LSTM possess stronger ability to capture temporal and spatial characteristic than traditional LSTM.
Why is it important?
we employ a deep learning (DL) model-the convolutional long short-term memory network (conyLSTM), to address the airport delay prediction in network structure. The spatiotemporal variables including flight delays of airport, air route congestion, airport throughput and flow control are input into an end-to-end learning architecture as a spatiotemporal sequence. The future flight delays in airport will be output by the model.