海南高尔夫球会收入时间序列模型选择及预测
ime Series Model Selection and Prediction of Golf Course Income in Hainan
高尔夫旅游对于建立海南国际旅游岛起着重要的作用，整理博鳌亚洲论坛国际会议中心高尔夫球会每月收入对数差分数据，BDS独立性检验、Box-Pierce、Box-Ljung独立性检验表明具有相关性。IAAFT代替数据非线性检验、White人工神经网络检验和Terasvirta人工神经网络非线性检验表明只存在线性相关性。利用自相关函数和偏相关函数以及AIC准则，选择周期13月一阶差分+ARMA(3,2)模型，极大似然估计参数，预留13个月数据相对误差在5%左右，表明模型选择准确可靠。最后预测2011年1月-2012年1月博鳌高尔夫球会月收入均值和95%置信区间。
Golf tourism plays an important role in construction of international island of Hainan. Logarithm differences of monthly golf course incomes in Boao Forum for Asia were collected. Serial dependence was examined by BDS, Box-Pierce and Box-Ljung independence tests. Only linear correlation was manifested using IAAFT surrogate data test, white and Terasvirta neural network test. Autocorrelation functions, partial correlation functions and AIC criterion were adopted to order identification. ARMA(3,2) model with 13 months period was the appreciated one. Parameters were estimated by maximum likelihood function. Relative errors of preserved datum were controlled almost within five percent. This certified our model's reliability. The mean and 95% confidence interval of Boao golf incomes from January, 2011 to January, 2012 were forecasted.
王洪礼、段永恒、李胜朋
旅游经济财政、金融
高尔夫旅游独立性检验非线性检验RMA
golf tourismindependence testnonlinearity testARMA
王洪礼,段永恒,李胜朋.海南高尔夫球会收入时间序列模型选择及预测[EB/OL].(2011-03-22)[2025-08-16].http://www.paper.edu.cn/releasepaper/content/201103-904.点此复制
评论