Research Of Flight Segment Volume Forecasting Based On Support Vector Machine

With competition becoming more and more intense in air transport market, the airline pay more attention on flight segment volume forecasting. For the airline, flight segment volume forecasting is not only relating to making important decision for the future development but also relating to making flight plan and aircraft formation plans. Therefore research of flight segment volume forecasting is necessary.This thesis analyzes forecasting methods and makes research of flight segment volume based on support vector machine. First, commonly used methods of flight segment volume forecasting are systematically elaborated, Then theory basis of support vector machine and its principle are introduced, support vector machine model is inferred, the support vector machines algorithm is studied. Model building method based on support vector machine is discussd, including choice of kernel function, choice of kernel model parameters, learning algorithm improvement and so on. Then forecasting models based on support vector machine are built and compared with neural network model. Experimental result indicated that, support vector machine used in flight segment volume is effective. Finally,forecasting method based on support vector machine applies in the system of "Flight Optimization & Revenue Maximization".