An Efficient Method to Detection Flight Delay Propagation based on Supplement GRU

As a result of the current flight delay, many other flights in the future are also delayed, and existing flight operations have been disrupted. The rate of aircraft delays and the consequent economic loss can be managed by civil aviation agencies with the help of analysis and forecasting of flight delay propagation in advance. Data on flight chains is notoriously difficult to compile due to the small number of possible data combinations. Traditional machine learning algorithms with a small sample size have been widely used in recent years for the investigation of the flight delay propagation problem. Preprocessing, feature selection, and model performance evaluation are all used in the suggested method. Data cleaning, data fusion, and normalization are used in the suggested method for preprocessing. IG is used for selecting features to analyze. The proposed method employs supplementary GRU for assessing the model's performance. When compared to two other methods, such as CNN and GRU, the proposed strategy performs exceptionally well.