Chapter 12

# Chapter 12 - Chapter 12 Multiple Regression 1 We can extend...

This preview shows pages 1–12. Sign up to view the full content.

1 Chapter 12 Multiple Regression

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
2 We can extend the method of least squares estimation to analyze the relationship the response variable has with more than 1 predictor variable. General Form: Method of least squares says that we are to minimize To minimize this, we would differentiate with respect to each b i , set each derivative equal to 0 and solve for each b i to get a fitted model. This is most easily accomplished via software (Minitab, SAS, SPSS) or the use of matrices. Fitted Model: i i i e y + + + + = k k i 1 1 0 X B X B B 2 1 1 0 2 ) ... ( ki k i i i x b x b b y e - - - - = k k 1 1 0 X B X B B y + + + =
3 12.2 Matrix Approach to Estimation Our model is Y = X Β +E Solving for Β, we get b=(X T X) -1 X T y where = = = - - - - - k k nk n n k n n n k k n n b b b b b x x x x x x x x x x x x X y y y y y 1 1 0 2 1 , 1 2 , 1 1 , 1 2 22 21 1 12 11 1 2 1 ... , ... 1 ... 1 ... ... ... ... ... ... 1 ... 1 , ... Most calculators have a matrix entry area so you can easily multiply the matrices to obtain the vector containing the estimates of B .

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
4 Example. Estimate the regression coefficients for the following data where x 1 =oil viscosity, x 2 =load and y=wear of a bearing: Y X 1 X 2 193 16 851 230 15.5 816 172 22 1058 91 43 1201 113 33 3571 125 40 1115 Identify the matrices y, X, and b. b =
5 y = X =

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
6 (X T X) = (X T X) -1 = (X T X) -1 X T = 17947 263626 8612 26362 25 . 5518 5 . 169 8612 5 . 169 6 - - - - - - - - - - - 7 6 4 6 4 992296281 . 1 551269664 . 5 291375582 . 1 551269664 . 5 0015247764 . 0351070108 . 291375582 . 1 0351070108 . 34375162 . 1 e e e e e - - - - - - - - - - - - - - - - - 4 4 4 5 5 5 290473094 . 1 0196943794 . 2044736453 . 991195448 . 3 0046129737 . 2758864128 . 285673704 . 1 0237912993 . 3209005076 . 048054426 . 4 261086144 . 5 841345929 . 4 0074351734 . 0160028127 . 0154347189 . 4348133892 . 6942602482 . 6721869283 . e e e e e e
7 b = (X T X) -1 X T y= So, our regression model is

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
8 Partitioning the Variation (ANOVA) We can use either matrices or statistical software, such as JMP, to find SS Total , SS Regression. SS Error . SS total = y[I-1(1’1) -1 1]y= SS Regression = y[X(X’X) -1 X’-1(1’1) -1 1]y= SS Error = y[I-X(X’X) -1 X’]y=SS Total -SS Regression 2 ) ( y y i - 2 ) ( y y i -
9 Source of Variation SS df MS F Regression k= Error N-(k+1)= Total N-1=

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Section 12.5 Hypothesis Testing Test for the significance of the regression model (F test) Tests for the model parameters (t test) 10
11 Hypothesis Test for the Model H 0 : B 1 =B 2 =…. = B k =0 H 1 : At least 1 B i is different from 0. ( This means

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 01/17/2011 for the course STAT 4714 at Virginia Tech.

### Page1 / 37

Chapter 12 - Chapter 12 Multiple Regression 1 We can extend...

This preview shows document pages 1 - 12. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online