7-5-8

Graphing Paired-Variable Statistical Data

Cubic Regression

Model Formula: y = a·x3 + b·x2 + c·x + d

a : cubic regression coefficient

b : quadratic regression coefficient

c : linear regression coefficient

d : regression constant term (y-intercept)

r2 : coefficient of determination

MSe : mean square error

 

1

 

n

• MSe =

 

 (yi – (a·xi3+ b·xi2+ c·xi +d ))2
n –

4

 

i=1

Quartic Regression

Model Formula: y = a·x4 + b·x3 + c·x2 + d·x + e

a : quartic regression coefficient

b : cubic regression coefficient

c : quadratic regression coefficient

d : linear regression coefficient

e : regression constant term (y-intercept)

r2 : coefficient of determination

MSe : mean square error

• MSe =

1

 

n

4

3

2

 

2

 

 

 

 

 

i=1

(yi – (a·xi
+ b·xi
+ c·xi

+ d·xi + e))

 

n –

5

 

20060301