旅游业趋势预测方法比较

旅游业对许多国家来说,是重要经济资源,带动的产业包括交通运输、住宿、餐饮、娱乐、零售等,对整体服务业与国家经济影响深远;但遇到重大事件冲击时,不但自身会受到严重伤害,还将波及相关产业,如何预测与规划未来旅游业的需求,已成为当务之急.本研究针对台湾主要客源地区(日本、港澳、美国、韩国、新加坡、马来西亚),借重人工神经网络的类型辨识与学习能力,加上警告信息变项,进行个别市场分析,来推测重大事件可能造成的冲击,以及主要境外观光客到达人数复苏的情形,期望能为旅游业的发展提供较精确的预测工具.分析结果证实:在指数平滑法、ARIMA和人工神经网络中,以人工神经网络预测效果最好.

Comparison Among Forecasting Approaches in Tourism Industry

As one of the most important economic resources among many countries, the tourism industry has a profound effect on overall aspects of service industries such as transportation, accommodation, catering, entertainment and retailing sectors. However, it is vulnerable when serious events happen to accurate demand forecasting and planning becomes a pressing need; this calls for a better solution. This study analyzes the tourist arrivals from each market among Japan, Hong Kong and Macao, U.S.A., Korea, Singapore, Malaysia for resource planning. To achieve the goal, an artificial neural network(ANN) is implemented with its prominent ability of learning and pattern recognition, which hopefully provides an accurate demand forecasting and benefits the whole service industry. The results reveal that ANN has the best performance among exponential smoothing, ARIMA and ANN methods.