You are on page 1of 5

1

1.1

1.2












2






F



3

1 b 1 1
= a + Y = , X =
y x y x

1 COD (y)
(x1)(x2)(x3)(x4) y

(1)

x1=[1.376, 1.375, 1.387, 1.401, 1.412, 1.428, 1.445, 1.477]


x2=[0.450, 0.475, 0.485, 0.500, 0.535, 0.545, 0.550, 0.575]

x3=[2.170 ,2.554, 2.676, 2.713, 2.823, 3.088, 3.122, 3.262]

x4=[0.8922, 1.1610 ,0.5346, 0.9589, 1.0239, 1.0499, 1.1065,


1.1387]

y=[5.19, 5.30, 5.605.826.00, 6.066.45 6.95]

(2)(.mat )

save data x1 x2 x3 x4 y

load data ()

(3)

x = [ones (8,1); x1 ' x2 ' x3 ' x4 ']

[bbintrrintstats] = regress(y,x)

b = (-16.528315.72062.0327-0.2106-0.1991)

stats = (0.990880.95300.0022)

y == -16.5283 + 15.7206xl + 2.0327x2 0.2106x3 + 0.1991x4

R2 = 0.9908F = 80.9530P = 0.0022

R2 R
1
F 80.9530 F0.05 (4,3) = 9.12

2 nlinfit

[beta,r,j] = nlinfit(xy'modelbeta0)

xy nm n
x n model m-beta0
beta r j Jacobian

[y,delta] = nlpredci(modelxbeta,r,j)

1 COD (y) logistic

a
y=
1 + be - ct

(1) m- mode1.m

function yhat=model(beta,t)

yhat=beta(1)(1+beta(2)*exp(-beta(3)*t))

(2)

t=18

load data y( data.mat y)

beta0=[50101]

(3)

[beta,r,j]=nlinfit(t,ymodelbeta0)

beta=(56.115710.40060.0445)

56.1157
y =
1 + 10.4006e -0.0445t

(4)

[yydelta] = nlprodei(modeltbeta,rj)

plot(tyk+tyy,r)

3
stepwise

stepwise(xyinmodelalpha)

x y nm nl
inmodel ()alpha,(
0.5)

stepwise plot
0
stepwise Table

(RMSE) (R-square)F P

eig()

[VD] = eig(R)

R X D R V

3 1

(1)

load data

x = [ x1 , x2 , x3 , x4 ]

(2)

R = corrcoef(x)

(3)

[VD] = eig(R)


0.3139 -0.7597 -0.1693 0.5438
-0.8143 0.0930 -0.1588 0.5505
v=
0.4860 0.6418 -0.2484 0.5387

0.0473 0.0485 0.9404 0.3332

0.0231
0 0 0
0 0.0601 0 0
d =
0 0 0.7306 0

0 0 0 3.1863

f1 = 0.5438xl+0.5505x2+0.5387x3+0.3332x4

3.18634 = 79.66

f2 = -0.1693xl-0.1588x2 0.2484x3 +0.9404x4

0.76064 = 19.12

f3 = -0.7597x1 + 0.0930x2 + 0.6418x3 + 0.0485x4

0.0601/4=1.5%

You might also like