24小时热门版块排行榜

返回列表

jiangwu8888

新虫 (正式写手)

应助: 4 (幼儿园)
金币: 6023.1
帖子: 818
在线: 258.9小时
虫号: 1051397

[交流] 【讨论】Matlab BP-ANN识别分类问题求教

各位大侠好，我最近在做BP的识别分类，用以下程序得到以下结果：
ans =
Columns 1 through 9
0.7603 0.1980 0.3132 0.7036 0.6818 0.2112 0.2669 0.7126 0.5661
0.0317 0.4560 0.6874 0.2286 0.2770 0.0488 0.7101 -0.2052 0.3679
-0.1144 0.4950 -0.0765 0.0134 -0.2668 0.7390 -0.1043 0.5205 0.0557
Columns 10 through 17
0.0680 0.1775 0.3187 0.0285 -0.5655 0.2369 0.6151 -0.0871
0.7512 0.4484 0.7357 0.7189 0.0631 0.0402 0.4088 0.3003
0.2376 0.1401 -0.2712 0.0661 0.9618 0.7444 -0.0919 0.7785
程序：
x1=[29.04 23.91 26.14 18.38 26.10 21.39 17.13 21.03 22.65 12.18 31.57 26.05 34.43 16.34 21.45 33.05 27.49 21.62 33.35 29.28 12.91 17.16 13.10 15.92 13.78 17.46 14.19 25.33 9.80 15.92;

37.21 22.25 21.06 21.97 30.63 25.93 19.69 17.42 27.75 14.44 17.09 22.45 18.45 14.52 18.66 20.49 16.43 22.51 25.05 17.61 8.75 12.72 11.69 10.94 14.62 15.59 12.1 16.21 11.1 17.34;

121.95 96.89 57.76 62.58 71.56 81.33 79.48 70.05 115.33 94.22 93.83 60.23 91.91 57.74 109.26 72.56 155.55 112.46 83.40 61.09 74.75 95.06 117.26 92.7 144.36 123.98 137.81 150.64 77.57 140.71;

184.18 155.75 139.04 119.43 155.14 126.67 158.80 133.84 201.27 105.86 111.81 130.59 172.08 131.27 159.63 145.26 141.63 165.62 147.58 147.59 125.34 138.05 159.25 125.1 151.8 172.59 188.17 231.6 114.98 181.71]

t=[1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;

0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0;

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1]

y=mapminmax(x1,0,1)
P=y
T=t

net_1=newff(minmax(P),[11,3],{'tansig','tansig'},'traingdm')

inputWeights=net_1.IW{1,1}
inputbias=net_1.b{1}
layerWeights=net_1.LW{2,1}
layerbias=net_1.b{2}

net_1.trainParam.show = 50;
net_1.trainParam.lr = 0.05;
net_1.trainParam.mc = 0.9;
net_1.trainParam.epochs = 1000;
net_1.trainParam.goal = 0.1;
[net_1,tr]=train(net_1,P,T);

A = sim(net_1,P);

E = T - A;

MSE=mse(E)

x2=[30.73 19.08 23.46 16.79 20.564 16.58 30.76 18.49 25.11 30.6 24.85 31.33 28.58 9.44 13.26 23.75 18.2;

24.37 14.77 18.29 20.20 23.57 27.90 19.55 18.06 16.92 17.90 13.57 21.54 16.65 12.93 11.56 21.21 14.55;

84.45 44.13 61.99 31.55 79.88 110.07 83.05 102.75 82.09 68.26 94.9 104.26 75.77 134.23 78.64 80.45 126.21;

206.40 111.29 122.22 131.34 137.81 177.25 130.69 189.92 180.90 112.98 175.16 136.71 116.4 135.26 152.08 152.08 179.77]

x=mapminmax(x2,0,1)

sim(net_1,x)

不知道结果怎么看？哪位大侠帮我分析下。（问题一）
结果如果不行，应该如何改进。（问题二）
希望能得到高人指点，谢谢

回复此楼