Improved grey prediction models for the trans-Pacific air passenger market

Abstract

The rapid economic growth of Asia-Pacific countries continues to result in faster travel growth in the trans-Pacific air passenger market. Grey theory is used to develop time series GM(1,1) models for forecasting total passenger and 10 country-pair passenger traffic flows in this market. The accumulated generating operation (AGO) is one of the most important characteristics of grey theory, and its main purpose is to reduce the randomness of data. The original GM(1,1) models are improved by using residual modifications with Markov-chain sign estimations. These models are shown to be more reliable by posterior checks and to yield more accurate prediction results than ARIMA and multiple regression models. The results indicate that the total number of air passengers in the trans-Pacific market will increase at an average annual growth rate of approximately 11% up to the year 2000.