基于神经网络和SARIMA模型的旅游需求探讨

拟对海南省旅游需求进行预测，采用旅游人数来度量旅游需求，收集相关部门数据，并通过分析旅游资源、环境、交通、费用和服务质量因素对旅游需求的影响，从而建立多元线性回归模型．在预测时，采用GM（1，1）得出各因素的预测值，然后通过神经网络进行海南省年旅游人数的预测，在对年内每月的旅游需求进行预测时，还考虑季节对旅游需求的影响，通过时间序列分析法，建立了SARIMA（3，1，2）（1，1，1）12模型，并进行了预测，结果表明，预测值符合实际人数．

Discussion on Tourism Demand-based on Neural Network and SARIMA Model

To predict the tourism demand of Hainan province, the number of tourist was used to determine the tourism demand, the related data were collected, and the effects of the tourism resources, the environment, transportation, cost and service quality factors on the tourism demand were analyzed, and a multiple linear re-gression model was established. The GM （ 1, 1 ） model was performed to obtain the predict value, and the neu-ral network model was used to predict the number of tourist each year of Hainan province. When the effects of season were taken into account, the SARIMA （3, 1,2） （ 1, 1, 1） 12 model was established for predicting. The results indicated that the predict number was keeping with the actual number.