ISBN: 978-981-14-4787-7 DOI: 10.18178/wcse.2020.06.010
Flight Delay Prediction Based on Elastic Neural Network
Abstract— Accurate prediction of flight delays is a difficult issue for airlines operation. Flight delays are affected by a variety of factors. It is difficult to accurately predict flight delay time from the perspective of traditional statistics. In order to reduce the data over-fitting, this study uses the genetic algorithm to select 21 related features. Then the regularized parameter L1 norm and L2 norm are introduced. Furthermore, the elastic neural network flight delay time prediction model is established to predict flight landing delay time. Finally, the accuracy within ±3 minutes tolerance is 83.954%, and the accuracy within ±5 minutes tolerance is 94.431%. The results show that the proposed model can improve the accuracy of flight delay prediction compared with the traditional simulation model.
Index Terms—Air traffic, flight delay prediction, genetic algorithm, feature selection, elastic neural network, Machine learning
Chaoyang Lu, Shuze Hang, Meize Dai
College of Civil Aviation, Nanjing University of Aeronautics and Astronautics, CHINA
Cite: Chaoyang Lu, Shuze Hang, Meize Dai, " Flight Delay Prediction Based on Elastic Neural Network " Proceedings of 2020 the 10th International Workshop on Computer Science and Engineering (WCSE 2020), pp. 57-61, Shanghai, China, 19-21 June, 2020.