Trip Generation Modeling Linear Regression - PowerPoint PPT Presentation

1 / 12
About This Presentation
Title:

Trip Generation Modeling Linear Regression

Description:

Trip Generation Modeling. Linear Regression. CE 573 Transportation Planning. Lecture 5 ... inflates with each addition of an independent variable. 12/21/09 ... – PowerPoint PPT presentation

Number of Views:939
Avg rating:3.0/5.0
Slides: 13
Provided by: Michae85
Category:

less

Transcript and Presenter's Notes

Title: Trip Generation Modeling Linear Regression


1
Trip Generation ModelingLinear Regression
  • CE 573 Transportation Planning
  • Lecture 5

2
Objectives
  • Regression review
  • Trip generation and regression

3
Linear Regression Review
  • Raw data
  • Dependent variable (Y)
  • Independent variable (X)
  • Generic model?

4
Linear Regression Review
  • Calibrating the model
  • Solving for b
  • Deviation in X?
  • Equation 4.6 for b?

5
Linear Regression Review
  • Calibrating the model
  • Solving for a
  • Average deviation in zero intercept model
    Equation 4.5 corrected

6
Linear Regression Review
  • Variance of the estimators

7
Linear Regression Review
  • Hypothesis tests for b
  • Ho b 0 H1 b ? 0
  • Ho rejected if 0 is not with in an appropriate
    confidence interval of the estimated parameter
  • test statistic Equation 4.8 and variance of the
    model residuals

8
Linear Regression Review
  • Hypothesis testing
  • t vs tcritical http//en.wikipedia.org/wiki/Stude
    nt27s_t-distribution
  • See t-distribution table

9
Multiple Regression
  • Multicollinearity Independent variables
    linearly related
  • Number of independent variables??
  • include if
  • strong theoretical backing
  • needed for policy decisions
  • Coefficient of determination (R2)
  • inflates with each addition of an independent
    variable

10
Trip Generation Modeling Using Regression
  • Aggregate (zonal)?
  • Disaggregate (household) ?
  • Zonal based multiple regression
  • Sample units?
  • Independent variables ?
  • Define the variables in terms of ___________
  • Household based regression
  • Sample units ?

11
Regression and Non-Linearity
  • Variable transformation
  • Y x2 relationship

12
Regression and Non-Linearity
  • Dummy variables
  • 0 car model ?aX
  • 1 car model ?bX
  • 2 car model ?cX
Write a Comment
User Comments (0)
About PowerShow.com