【答案】应助回帖
★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★
感谢参与,应助指数 +1
fegg7502: 金币+1, 鼓励交流 2013-05-18 14:37:38
jadetsai88(fegg7502代发): 金币+5 2013-05-18 14:38:16
jadetsai88: 金币+45, ★★★★★最佳答案 2013-05-18 20:09:23
列出10个相关系数R^2最高的拟合方程,这些方程都能到0.9999高相关度,取一个形式简单的即可,如下:
1.Function: y = p1+p2*Ln(x)+p3*(Ln(x))^2+p4*(Ln(x))^3+p5*(Ln(x))^4+p6*(Ln(x))^5+p7*(Ln(x))^6+p8*(Ln(x))^7+p9*(Ln(x))^8+p10*(Ln(x))^9+p11*(Ln(x))^10
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0816570704464187
Sum of Square Error (SSE): 2.80717628178828
Correlation Coef. (R): 0.999952519401421
R-Square: 0.99990504105725
Determination Coef. (DC): 0.99990504105704
Parameters Name Parameter Value
=============== ===============
p1 104882.177175265
p2 -170072.882367486
p3 41858.0748248411
p4 5855.07829941347
p5 -2064.98454070029
p6 -57.5187050922973
p7 -1.65547534829854
p8 10.0923903851733
p9 -0.051277245249696
p10 -0.244342059686163
p11 0.0176782129053633
2.Function: y = p1+p2*x+p3/Ln(x)+p4/x^0.5+p5/x^2
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0818929502757209
Sum of Square Error (SSE): 2.82341768334678
Correlation Coef. (R): 0.99995224468758
R-Square: 0.999904491655731
Determination Coef. (DC): 0.999904491655731
Parameters Name Parameter Value
=============== ===============
p1 139056.408025523
p2 -8.53609748799606
p3 -1168509.96880985
p4 1215831.33120729
p5 -74638263.833034
3.Function: y = p1+p2*x+p3/Ln(x)+p4/x^0.5+p5*Ln(x)/x^2
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0819040674378914
Sum of Square Error (SSE): 2.82418430666855
Correlation Coef. (R): 0.999952231720598
R-Square: 0.999904465723005
Determination Coef. (DC): 0.999904465723002
Parameters Name Parameter Value
=============== ===============
p1 158206.095651443
p2 -9.24552847545844
p3 -1337991.6917642
p4 1409520.13162155
p5 -19485218.4487566
4.Function: y = p1+p2*x+p3*(Ln(x))^2+p4*Ln(x)/x+p5*Ln(x)/x^2
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.103069263313438
Sum of Square Error (SSE): 4.47239794982936
Correlation Coef. (R): 0.999952224862649
R-Square: 0.999904452007761
Determination Coef. (DC): 0.999848711253166
Parameters Name Parameter Value
=============== ===============
p1 -23856.4123261917
p2 -9.5166822779122
p3 630.611936026086
p4 457692.303453987
p5 -23822706.0307153
5.Function: y = p1+p2*x+p3*x^0.5*Ln(x)+p4*Ln(x)/x+p5/x^2
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.088522095717737
Sum of Square Error (SSE): 3.29902396213954
Correlation Coef. (R): 0.999952224257904
R-Square: 0.99990445079833
Determination Coef. (DC): 0.999888403222029
Parameters Name Parameter Value
=============== ===============
p1 -8499.42236763435
p2 -17.713721636907
p3 111.174591525206
p4 244694.92650816
p5 -74289419.0906333
6.Function: y = p1+p2*x+p3*x^0.5+p4/x+p5*Ln(x)/x^2
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0938796382387353
Sum of Square Error (SSE): 3.71043570632687
Correlation Coef. (R): 0.999952220269342
R-Square: 0.999904442821587
Determination Coef. (DC): 0.999874486310361
Parameters Name Parameter Value
=============== ===============
p1 -12262.6182444782
p2 -14.3627362170755
p3 783.568521217771
p4 1721369.41889051
p5 -27483015.6528191
7.Function: y = p1+p2*Ln(x)+p3/Ln(x)+p4*(Ln(x))^2+p5/(Ln(x))^2+p6*(Ln(x))^3+p7/(Ln(x))^3+p8*(Ln(x))^4+p9/(Ln(x))^4+p10*(Ln(x))^5
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0819153678857709
Sum of Square Error (SSE): 2.82496367575756
Correlation Coef. (R): 0.999952218538026
R-Square: 0.99990443935912
Determination Coef. (DC): 0.99990443935912
Parameters Name Parameter Value
=============== ===============
p1 -643037.084034645
p2 38.9986995014769
p3 4524064.61458851
p4 -0.520688702667894
p5 -8854121.10254966
p6 1312.95358746647
p7 215.420054660352
p8 -114.696108296773
p9 222.271294032126
p10 0.0952928226364861
8.Function: y = p1+p2*Ln(x)+p3/Ln(x)+p4*(Ln(x))^2+p5/(Ln(x))^2+p6*(Ln(x))^3+p7/(Ln(x))^3+p8*(Ln(x))^4
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0819154627883818
Sum of Square Error (SSE): 2.82497022145444
Correlation Coef. (R): 0.999952218427348
R-Square: 0.999904439137775
Determination Coef. (DC): 0.999904439137697
Parameters Name Parameter Value
=============== ===============
p1 -297.687216527561
p2 -7518.71078893706
p3 -50206.0976817148
p4 -8101.57288160546
p5 4825074.89629612
p6 2150.26094280003
p7 -15466014.2875334
p8 -139.286181651244
9.Function: y = p1+p2*x+p3*(Ln(x))^2+p4*Ln(x)/x+p5/x^2
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0937298793755589
Sum of Square Error (SSE): 3.69860721114562
Correlation Coef. (R): 0.999952213783732
R-Square: 0.999904429850987
Determination Coef. (DC): 0.999874886435357
Parameters Name Parameter Value
=============== ===============
p1 -19039.1963454338
p2 -8.28803860922449
p3 520.04483941951
p4 343609.017628046
p5 -83652096.5281173
10.Function: y = p1+p2*x+p3*x^0.5+p4*Ln(x)/x+p5*Ln(x)/x^2
Algorithms: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
Root of Mean Square Error (RMSE): 0.0955640457041217
Sum of Square Error (SSE): 3.84477695599391
Correlation Coef. (R): 0.999952213440392
R-Square: 0.99990442916434
Determination Coef. (DC): 0.999869941920631
Parameters Name Parameter Value
=============== ===============
p1 -15081.2801117689
p2 -15.7387786911996
p3 891.516177188911
p4 352116.987909116
p5 -21092486.2844489
![]()
附图1.jpg
|