基于混合编码和长短时记忆网络的机场延误预测方法

为充分利用机场延误状态信息的时间相关性，提高机场延误预测精度，提出一种基于混合编码和长短时记忆网络（Long Sort Term Memory, LSTM）的机场延误预测方法。该方法首先将机场信息、航班信息和气象信息进行数据预处理，得到机场延误数据；然后，利用LSTM网络对机场延误数据进行特征提取；最后，构建Softmax分类器对机场延误分类预测。实验结果表明，本文基于机场延误数据在数据预处理阶段提出的混合编码方法，可使预测准确率提高约5%。同时，利用LSTM网络来提取数据的时间相关特征信息，网络模型的预测准确率最终可达94.01%。并且由不同机场数据对网络的普适性分析结果表明，该算法更适合于原始数据量大的中大型枢纽机场。 

Airport Delay Prediction Method Based on Hybrid Coding and LSTM Network

In order to make full use of the time correlation of airport delay status information and improve the accuracy of airport delay prediction, an airport delay prediction method based on hybrid coding and Long Short Term Memory（LSTM） network is proposed. The method firstly preprocesses the airport's information, meteorological information and flight information to obtain the airport delay data. Then, the LSTM network is used to extract the feature of the airport delay data. Finally, the Softmax classifier is constructed to predict the classification of airport delays. The experiments results show that the hybrid coding method proposed in the data preprocessing stage based on airport delay data can improve the prediction accuracy by about 5%. Moreover, when the LSTM network is used to extract the time-related feature information of the data, the prediction accuracy of the LSTM network model is up to 94.01%. And the analysis of the universality of the network using different airport data shows that the algorithm is more suitable for medium and large hub airports with large amount of original data.