%%
%對程序的說明%
%做生成數（使用0.1做因子）之后13年的數據按統計方法標準化后輸入訓練仿真對比結果。這個過程相當于插值計算%
%輸出結果請最后五句r1rlplotyy（ktokt）plotyy（kakf）%

%這樣做的缺點是不能檢驗預測效果的好壞還有一點我對網絡初始化的過程沒有把握%

clear
%?數據輸入
inv=[189.6??244.2??328.5??405.3??487.5??568.6??669.9??801.6??785.9??859.2??979.7??1086.3??1241];
d1=[15.3??16.6??17.2??22.8??37.6??30.8??38.9??50.8??58.2??41.2??55.8??71.7??82.3];
d2=[117.7??140.7??177.8??209.8??247.7??290.8??343.2??337.1??308.1??353.6??350.1??367.2??415.7];
d3=[56.6??86.9??133.5??172.6??202.2??247??287.8??413.7??419.6??464.4??573.8??647.4??742.1];
f=[822.3??959.7??1198.4??1607.5??1997.6??2380.9??2683.8??2798.9??2897.4??3253??3561??3882?4433];
k=1991:1:2003;
%?使用13年數據訓練得到網絡仿真計算相對誤差
%對原始數據初始化的過程為t（1）=f（1）;??t（i）=t（i-1）+0.1*f（i）（i=2:13）
%?累加?p1（1）=d1（1）;for?i=2:13?p1（i）=0.1*d1（i）+p1（

?屬性????????????大小?????日期????時間???名稱
-----------?---------??----------?-----??----

?????文件???????2474??2005-05-22?15:06??bp-matlab\bp源程序\bp.m

?????文件???????1988??2005-06-13?21:54??bp-matlab\bp源程序\bp2.m

?????文件??????28792??2004-08-11?20:11??bp-matlab\bp源程序\sourse\ANN.mat

?????文件???????8442??2004-08-12?10:20??bp-matlab\bp源程序\sourse\AnnData.txt

?????文件???????1388??2004-08-12?10:32??bp-matlab\bp源程序\sourse\Predict.asv

?????文件???????1371??2004-08-12?10:27??bp-matlab\bp源程序\sourse\Predict.m

?????文件????????273??2004-08-12?10:34??bp-matlab\bp源程序\sourse\說明.txt

????..AD...?????????0??2006-03-18?19:39??bp-matlab\bp源程序\sourse

?????目錄??????????0??2006-03-18?19:38??bp-matlab\bp源程序

?????目錄??????????0??2010-04-14?10:15??bp-matlab

-----------?---------??----------?-----??----

????????????????44728????????????????????10