Abstract：Airline ticket purchasing is convenient with the development of Internet,and airfare forecast can help passengers or travel agents explore the market and make sophisticated decisions.However,airline companies are using complicated strategies,Being effected by multiple dynamically factors,the real-time price of tickets changes rapidly and has the properties of tendency,randomness and large floating,so we can′t find out an accurate relationship between the history and the future.In this paper,we proposed an airfare forecast model based on fuzzy time series.We analyze the lowest price of 36 airlines from October 2014 to April 2015 and use MAPE as the indicator of prediction results.A comparison was made with three classical time series algorithms：AR,moving average and exponential smoothing,and prove that the model of fuzzy time series can increase the accuracy of the forecast,especially when the price changes significantly and the performance of each line are less floating.As a result,airfare forecast can provide passengers a pretty reliable decision support.
华逸群,曹健. 机票价格预测的模糊时间序列方法[J]. 小型微型计算机系统, 2016, 37(11): 2547-2551.
HUA Yi-qun,CAO Jian. Airfare Forecast Based on Fuzzy Time Series. Journal of Chinese Computer Systems, 2016, 37(11): 2547-2551.
2016, Vol. 37 
