R 2 =0.9840 较高,且都是一次项,计算简便。缺点自相关性未能判定,对 1981 年之后的数据预测值逐渐产生偏差。六、模型的推广与改进从前文的分析来看,我们认为 1981 年后,物价、通货膨胀、人民币币值等因素极大地影响了我们的预测数据,若补充上物价衡量指数,消费水平指数( CPI )等数据,模型的拟合度应该会更高。七、附件模型( 1 )的 MATLAB 程序: clear,clc % 清屏 x1=[598,586,707,737,825,837,1028,1114,1079,757,677,779,943,1152,1322,1249,1187,1372,163 8,1780,1833,1978,1993,2121,2052,2189,2475,2702,2791]'; x2=[349,455,520,558,715,798,1235,1681,1870,1156,964,1046,1250,1581,1911,1647,1565,2101, 2747,3156,3365,3684,3696,4254,4309,4925,5590,6065,6592]'; x3=[461,475,491,529,556,575,598,509,444,434,461,514,584,632,687,697,680,688,767,790,789,8 55,891,932,955,971,1058,1150,1194]'; x4=[57482,58796,60266,61465,62828,64653,65994,67207,66207,65859,67295,69172,70499,725 38,74542,76368,78534,80671,82992,85229,87177,89211,90859,92421,93717,94974,96259,9754 2,98705]';
全文预览
