请教关于 matlab多元非线性逐步回归 相关的问题数据如下：y x1 x2 x3 x48.24-0.17573-0.0061019-0.050177-0.02027610.3-0.0826270.0055

请教关于 matlab多元非线性逐步回归 相关的问题数据如下：y x1 x2 x3 x48.24-0.17573-0.0061019-0.050177-0.02027610.3-0.0826270.0055
请教关于 matlab多元非线性逐步回归 相关的问题
数据如下：
y x1 x2 x3 x4
8.24-0.17573-0.0061019-0.050177-0.020276
10.3-0.0826270.0055711-0.03733-0.017611
9.360.00117880.015542-0.025828-0.015077
11.40.0756860.02381-0.015671-0.012675
12.360.140890.030377-0.0068601-0.010404
15.790.196810.0352410.00060578-0.0082634
16.210.243420.0384030.0067264-0.0062545
9.980.280730.0398630.011502-0.0043768
13.710.308750.0396210.014932-0.0026303
15.790.327460.0376760.017017-0.001015
19.950.336880.034030.0177560.00046912
12.380.353040.0205880.0188170.0030883
17.090.399040.00418570.014570.0024553
11.60.3202-2.88E-05-0.000641110.0091938
17.160.27621-0.0070493-0.00504590.01228
13.360.24993-0.016256-0.00990090.014044
4.520.2059-0.032168-0.0118340.011766
9.310.16562-0.038794-0.0213890.013523
4.820.050641-0.024166-0.0221920.012703
4.470.018767-0.021508-0.0224630.0071036
13.01-0.082144-0.021609-0.0141820.0044351
9.7-0.19178-0.014396-0.00700380.00034502
5.22-0.294730.00072856-0.00654930.0020938
16.6-0.28953-0.00633170.006708-0.0048055
11.57-0.322970.00690980.01876-0.0093256
13.76-0.284930.000921550.027494-0.012509
8.88-0.24188-0.00784880.035687-0.014758
15.2-0.19098-0.0215610.029592-0.0086577
13.79-0.24445-0.00711070.0210750.0050022
20.63-0.19383-0.0174550.0226580.0065334
14.02-0.12704-0.0331390.0249680.0052232
17.5-0.079408-0.0441640.0222680.0034713
14.43-0.079494-0.0364780.0107940.0052206
10.83-0.081988-0.027189-0.000543840.0070777
16.38-0.014893-0.035732-0.00130270.0065977
14.280.069564-0.037314-0.0021215-0.00021508
11.120.086443-0.0366250.0014271-0.0062919
17.170.087585-0.028083-0.00078778-0.0087206
15.760.081179-0.023345-0.0025491-0.0083321
13.440.064229-0.017348-0.0040484-0.0073553
9.910.036734-0.01009-0.0052859-0.0057901
14.93-0.0013051-0.0015727-0.0062615-0.0036366
14.96-0.0498890.0082048-0.0069751-0.00089464
13.12-0.109020.019242-0.00742690.0024357
15.93-0.178690.03154-0.00761680.0063544
10.16-0.258910.045098-0.00754470.010861
10.37-0.349670.059916-0.00721080.015957
12.81-0.450980.075993-0.00661490.021641
如何用非线性逐步回归进行拟合,最好给出代码和详细的解释,谢谢!

请教关于 matlab多元非线性逐步回归 相关的问题数据如下：y x1 x2 x3 x48.24-0.17573-0.0061019-0.050177-0.02027610.3-0.0826270.0055
T=[8.24\x09-0.17573\x09\x09-0.0061019\x09-0.050177\x09-0.020276;
10.3\x09-0.082627\x090.0055711\x09-0.03733\x09\x09-0.017611;
9.36\x090.0011788\x090.015542\x09\x09-0.025828\x09-0.015077;
11.4\x090.075686\x09\x090.02381\x09\x09-0.015671\x09-0.012675;
12.36\x090.14089\x09\x090.030377\x09\x09-0.0068601\x09-0.010404;
15.79\x090.19681\x09\x090.035241\x09\x090.00060578\x09-0.0082634;
16.21\x090.24342\x09\x090.038403\x09\x090.0067264\x09-0.0062545;
9.98\x090.28073\x09\x090.039863\x09\x090.011502\x09\x09-0.0043768;
13.71\x090.30875\x09\x090.039621\x09\x090.014932\x09\x09-0.0026303;
15.79\x090.32746\x09\x090.037676\x09\x090.017017\x09\x09-0.001015;
19.95\x090.33688\x09\x090.03403\x09\x090.017756\x09\x090.00046912;
12.38\x090.35304\x09\x090.020588\x09\x090.018817\x09\x090.0030883;
17.09\x090.39904\x09\x090.0041857\x090.01457\x09\x090.0024553;
11.6\x090.3202\x09\x09-2.88E-05\x09\x09-0.00064111\x090.0091938;
17.16\x090.27621\x09\x09-0.0070493\x09-0.0050459\x090.01228;
13.36\x090.24993\x09\x09-0.016256\x09-0.0099009\x090.014044;
4.52\x090.2059\x09\x09-0.032168\x09-0.011834\x090.011766;
9.31\x090.16562\x09\x09-0.038794\x09-0.021389\x090.013523;
4.82\x090.050641\x09\x09-0.024166\x09-0.022192\x090.012703;
4.47\x090.018767\x09\x09-0.021508\x09-0.022463\x090.0071036;
13.01\x09-0.082144\x09-0.021609\x09-0.014182\x090.0044351;
9.7\x09-0.19178\x09\x09-0.014396\x09-0.0070038\x090.00034502;
5.22\x09-0.29473\x09\x090.00072856\x09-0.0065493\x090.0020938;
16.6\x09-0.28953\x09\x09-0.0063317\x090.006708\x09\x09-0.0048055;
11.57\x09-0.32297\x09\x090.0069098\x090.01876\x09\x09-0.0093256;
13.76\x09-0.28493\x09\x090.00092155\x090.027494\x09\x09-0.012509;
8.88\x09-0.24188\x09\x09-0.0078488\x090.035687\x09\x09-0.014758;
15.2\x09-0.19098\x09\x09-0.021561\x090.029592\x09\x09-0.0086577;
13.79\x09-0.24445\x09\x09-0.0071107\x090.021075\x09\x090.0050022;
20.63\x09-0.19383\x09\x09-0.017455\x090.022658\x09\x090.0065334;
14.02\x09-0.12704\x09\x09-0.033139\x090.024968\x09\x090.0052232;
17.5\x09-0.079408\x09-0.044164\x090.022268\x09\x090.0034713;
14.43\x09-0.079494\x09-0.036478\x090.010794\x09\x090.0052206;
10.83\x09-0.081988\x09-0.027189\x09-0.00054384\x090.0070777;
16.38\x09-0.014893\x09-0.035732\x09-0.0013027\x090.0065977;
14.28\x090.069564\x09\x09-0.037314\x09-0.0021215\x09-0.00021508;
11.12\x090.086443\x09\x09-0.036625\x090.0014271\x09-0.0062919;
17.17\x090.087585\x09\x09-0.028083\x09-0.00078778\x09-0.0087206;
15.76\x090.081179\x09\x09-0.023345\x09-0.0025491\x09-0.0083321;
13.44\x090.064229\x09\x09-0.017348\x09-0.0040484\x09-0.0073553;
9.91\x090.036734\x09\x09-0.01009\x09\x09-0.0052859\x09-0.0057901;
14.93\x09-0.0013051\x09-0.0015727\x09-0.0062615\x09-0.0036366;
14.96\x09-0.049889\x090.0082048\x09-0.0069751\x09-0.00089464;
13.12\x09-0.10902\x09\x090.019242\x09\x09-0.0074269\x090.0024357;
15.93\x09-0.17869\x09\x090.03154\x09\x09-0.0076168\x090.0063544;
10.16\x09-0.25891\x09\x090.045098\x09\x09-0.0075447\x090.010861;
10.37\x09-0.34967\x09\x090.059916\x09\x09-0.0072108\x090.015957;
12.81\x09-0.45098\x09\x090.075993\x09\x09-0.0066149\x090.021641;];
p=T(:,2:5)';t=T(:,1)';
[pn,minp,maxp,tn,mint,maxt]=premnmx(p,t); %将数据归一化
NodeNum1 =20; % 隐层第一层节点数
NodeNum2=40; % 隐层第二层节点数
TypeNum = 1; % 输出维数
TF1 = 'tansig';TF2 = 'tansig'; TF3 = 'tansig';
net=newff(minmax(pn),[NodeNum1,NodeNum2,TypeNum],...
{TF1 TF2 TF3},'traingdx');%网络创建%traingdm
net.trainParam.show=50;
net.trainParam.epochs=50000;%训练次数设置
net.trainParam.goal=1e-5;%训练目标设置
PL.lr=0.01;
net=train(net,pn,tn);
an=sim(net,pn);
[a]=postmnmx(an,mint,maxt); %数据的反归一化
[m,b,r]=postreg(a,t)
save mynetdate net
y=an';
plot(1:length(t),t,'*',1:length(t),a,'o');
title('o表示预测值 AND *表示实际值')
grid on
error=t-a;
figure
plot(1:length(error),error,'-.')
title('误差变化图')
grid on
这种几个参数的抽象函数,人工神经网络是最佳处理方法.运行的结果还行.

这得多牛才能答得出啊。。
你还是去专业科研论坛去问吧。。比如小木虫什么的