Estimation of flight delay using weighted Spline combined with ARIMA model

Adding to societal changes today, are the miscellaneous big data have been produced in different fields. Coupled with these data is the appearance of data risk management and data mining. Admittedly, to predict future trend by using these data is conducive to make everything more efficient and easy. This paper develops a new prediction model of flight departure delay. By studying the main factors lead to flight delay, the paper takes weather, holiday influences and hourly pattern as variables of the mixed function by combining smoothing Spline function with ARIMA models. We optimized and simulated with 3 years of data from American Airline. By utilizing our model can predict delays of each flight on a specific day and hour. The result demonstrates the goodness of fit.