IOE366-Ch12-Simple+Regression

# IOE366-Ch12-Simple+Regression - Ch 12 Simple Linear...

This preview shows pages 1–6. Sign up to view the full content.

Ch 12: Simple Linear Regression and Correlation 12.1 Simple Linear Regression Model 12.2 Estimation of Model Parameters 12.3 Inference about the Slope 12.4 Inference about Future Y and E(Y) 12.5 Correlation 1 Ch 12 - Learning Objectives ± Use SLR for empirical model building nderstand Method of Least Squares ± Understand Method of Least Squares ± Analyze residuals to determine aptness of odels models ± Test hypotheses and construct confidence intervals ± Use SLR models to predict ± Apply correlation models 2 Regression Analysis and Empirical Models Regression Analysis is statistical technique for modeling and ± a statistical technique for modeling and investigating the relationship between two r more variables or more variables. n mpirical Model An Empirical Model is ± a statistical model based on observations ther than theory 3 rather than theory. Scatterplots ± Graphically show empirical relationships etween two variables between two variables ± Example: Torque vs Pressure 4

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

View Full Document
5 http://www.enerpac.com/html/products/industrial_tools/Bolting/Cat_pages/HXD_torqpress_325US.pdf Air Pressure Torque i 120.0 psi Nm 50 66.4 52 69.5 4 0 4 100.0 110.0 ue 54 70.4 56 71.9 58 75.4 60 82.8 80.0 90.0 Torq u 62 79.6 64 89.7 66 93.9 60.0 70.0 0 0 0 0 0 68 93.3 70 97.3 72 100.8 4 03 9 40 50 60 70 80 Air Pressure 74 103.9 76 103.4 78 113.6 80 108.8 Y: Torque X: Air Pressure Other variables? What information does one variable (X) provide about another (Y) ? Y is the Dependent Variable The variable one predicts (e.g. Torque) is the Independent Variable X is the Independent Variable Predictor Variable Explanatory Variable Want to estimate the best linear model describing the relationship between Torque and Air Pressure 7 pq 12.1 Simple Linear Regression Model Equation of a Straight Line with Error: 01 Y x β βε =+ + Y = dependent variable, random variable. = independent variable, nothing random about it. , = intercept and slope, unknown constants. 0 1 ε = random error term, ssumption: normally distributed with 8 Assumption: ε is normally distributed with E( ε )=0 and V( ε )= σ 2
Model has Two Components: ) Deterministic 1) Deterministic 2) Random 2 1 Assume () ~ N (0 ,) y x εβ β σ =− + 9 01 ii i ββ Simple Linear Regression Model the mean response of Y for a given X with inimal error minimal error. rror Terms Observed Value Error Terms + * Y i * Fit Line Y i * 10 * Simple Linear Regression We estimate the line with ˆ Y ’s: ˆˆ ˆ =+ (fitted value) and estimate the errors with ) + ey (residuals) 11 Criterion for Choosing “Best Fit” Line Error Terms: distance from fit line (predicted values) to observed data point (observed values) Criterion: 12 Called “Least Squares” or “Ordinary Least Squares”

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

View Full Document
Least Squares Criterion 22 01 Minimize ( ) nn ii y x εβ β = ∑∑ 11 == ince we don't know we need to Since we don t know , ββ estimate them using ˆˆ , 2 13 Minimize ( –– ) ey = Solving 2 Equations … 2 n e 1 0 2( ) 0 ˆ i yx ∂β = = =• = 2 1 1 ) 0 ˆ xy = = = 14 Solution () xx yy −− 1 1 2 ˆ ) = = 1 = 15 = Computational Formula, for 1 ˆ S = 1 ( )( ) / Sx =− ( ) 2 2 / 16
Example #1 : MINITAB Regression Output Torque (Nm) = -14.2456 + 1.58522 Air Pressure Regression Plot 115 S = 2.68025 R-Sq = 97.1 %

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 03/17/2010 for the course IOE 366 taught by Professor Garyherrin during the Winter '10 term at University of Michigan-Dearborn.

### Page1 / 20

IOE366-Ch12-Simple+Regression - Ch 12 Simple Linear...

This preview shows document pages 1 - 6. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online