least squares approximation: a linear algebra technique

13
Least Squares Approximation: A Linear Algebra Technique By Larry Wong and James Sfregola

Upload: gerodi

Post on 06-Jan-2016

39 views

Category:

Documents


2 download

DESCRIPTION

Least Squares Approximation: A Linear Algebra Technique. By Larry Wong and James Sfregola. Soo……You have a bunch of Data. What can we do with this? Use Least Squares Approximation (LSE) to estimate the relationship of the dependent variable and the independent variable(s). - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Least Squares Approximation: A Linear Algebra Technique

Least Squares Approximation: A Linear Algebra Technique

By Larry Wong and James Sfregola

Page 2: Least Squares Approximation: A Linear Algebra Technique

Soo……You have a bunch of Data.

Attendence Big 4 Derby League Position per capita incomeadjusted

population

39,968 1 0 11 35000 1582564

21,852 1 0 10 36725 137471

24,409 1 0 7 36725 371847

26,770 1 0 19 43600 852013

41,917 1 1 2 43600 1351187

37,162 1 0 6 36725 623160

24,510 1 1 16 43600 724121

43,958 1 1 3 36725 738873

40,699 1 0 14 36725 630043

75,595 1 1 1 36725 1194415 What can we do with this?

Use Least Squares Approximation (LSE) to estimate the relationship of the dependent variable and the independent variable(s).

Page 3: Least Squares Approximation: A Linear Algebra Technique

What is Least Squares?What is Least Squares?

• Approximates solutions for inconsistent Approximates solutions for inconsistent over determined systems over determined systems – What does inconsistent mean?What does inconsistent mean?

– Over determined, what’s that?Over determined, what’s that?

• In other words LSE allows you to describe In other words LSE allows you to describe a model as being represented by the line a model as being represented by the line of best fit.of best fit.

• Where the “best fit” line, curve or polynomial is Where the “best fit” line, curve or polynomial is created from the approximations.created from the approximations.

Page 4: Least Squares Approximation: A Linear Algebra Technique

Least Squares it’s Actually Pretty

Useful Least Squares Approximations are

highly relevant in various fields Really, how so?

Still don’t believe us? Fine we’ll give examples…but not now.

Page 5: Least Squares Approximation: A Linear Algebra Technique

So How Does it Estimate a Solution?

Least Squares Approximation estimates a best fit solution for a system by Minimizing the magnitude error

vector, e.

Estimating the coefficients of the best fit equation

Page 6: Least Squares Approximation: A Linear Algebra Technique

How Do We Ensure Positive Error?

Least Squares– Eliminates negative terms– More susceptible to outlying data

Least Absolute Difference– Difficult to work with in linear algebra

Page 7: Least Squares Approximation: A Linear Algebra Technique

`Petal L.`

`Petal

W.`

1 2 3 4 5 6 7

0.51.0

1.52.0

2.5

Page 8: Least Squares Approximation: A Linear Algebra Technique

The Least Squares Theorem For an over determined inconsistent

system Ax=b: This is easily solved using inverse and

transpose multiplication

the resulting vector x can be determined in the form x=(AT A) -1 AT b

where x will correspond to the coefficients in your linear or polynomial expression.

Page 9: Least Squares Approximation: A Linear Algebra Technique

How to Approximate the Error How to Approximate the Error Term?Term?

• ee = = bb-A-Axx , where , where e e is the error termis the error term

• Where e1= the Where e1= the distance from distance from our data point to our data point to the best fit the best fit approximation approximation given by the given by the orthogonal orthogonal projectionprojection

Page 10: Least Squares Approximation: A Linear Algebra Technique

Application to Economics

Econometrics utilizes Ordinary Least Squares Approximations (OLS) Similar to LSE except it follows 7

classical assumptions Estimated Regression Equation

Page 11: Least Squares Approximation: A Linear Algebra Technique

-30000

-20000

-10000

0

10000

20000

30000

0

20000

40000

60000

80000

50 100 150 200 250 300 350

Residual Actual Fitted

Page 12: Least Squares Approximation: A Linear Algebra Technique
Page 13: Least Squares Approximation: A Linear Algebra Technique

BibliographyBibliography

• Poole, David Poole, David Linear AlgebraLinear Algebra: A : A Modern Introduction. Canada: Modern Introduction. Canada: Thomson Brooks/Cole, 2006. Thomson Brooks/Cole, 2006.

• Studenmund, A.H. Using Studenmund, A.H. Using Econometrics: A Practical Guide. New Econometrics: A Practical Guide. New York: Pearson Education, 2006. York: Pearson Education, 2006.