Linear Regression Andy Jacobson July 2006 Statistical Anecdotes: Do hospitals make you sick? Student’s story Etymology of “regression”

Linear RegressionLinear Regression

Andy Jacobson

July 2006

Statistical Anecdotes:

Do hospitals make you sick?

Student’s story

Etymology of “regression”

Linear RegressionLinear Regression

Andy Jacobson

July 2006

Statistical Anecdotes:

Do hospitals make you sick?

Student’s story

Etymology of “regression”

OutlineOutline

1. Discussion of yesterday’s exercise2. The mathematics of regression3. Solution of the normal equations4. Probability and likelihood

5. Sample exercise: Mauna Loa CO2

6. Sample exercise: TransCom3 inversion

http://www.aos.princeton.edu/WWWPUBLIC/http://www.aos.princeton.edu/WWWPUBLIC/sara/statistics_course/andy/R/sara/statistics_course/andy/R/

corr_exer.r 18 July practical

mauna_loa.r Today’s first example

transcom3.r Today’s second example

dot-Rprofile Rename to ~/.Rprofile (i.e., home dir)

hclimate.indices.r Get SOI, NAO, PDO, etc. from CDC

cov2cor.r Convert covariance to correlation

ferret.palette.r Use nice ferret color palettes

geo.axes.r Format degree symbols, etc., for maps

load.ncdf.r Quickly load a whole netCDF file

svd.invert.r Multiple linear regression using SVD

mat4.r Read and write Matlab .mat files (v4 only)

svd_invert.m Multiple linear regression using SVD (Matlab)

atm0_m1.mat Data for the TransCom3 example

R-intro.pdf Basic R documentation

faraway_pra_book.pdf Julian Faraway’s “Practical Regression and ANOVA in R” book

Multiple Linear RegressionMultiple Linear Regression Data

ParametersBasis Set

Basis FunctionsBasis Functions

“Design matrix” A gives values of each basis function at each observation location.

Basis Functions

Observations

Note that one column of (e.g., ai1) may be all ones, to represent the “intercept”.

From the Cost Function From the Cost Function to the Normal Equationsto the Normal Equations

“Least squares” optimization minimizes sum of squared residuals (misfits to data). For the time being, we assume that the residuals are IID:

Expanding terms:

Cost is minimized when derivative w.r.t. x vanishes:

Rearranging:

Optimal parameter values (note that ATA must be

invertible):

x-hat is BLUE

BLUE = Best Linear Unbiased Estimate(not shown here: “best”)

Practical Solution of Normal Equations using SVDPractical Solution of Normal Equations using SVD

If we could pre-multiply the forward equation by A-1, the “pseudo-inverse” of A, we could get our answer directly:

For every M x N matrix A, there exists a singular value

decomposition (SVD):

U is M x MS is N x NV is N x N

S is diagonal and contains the Singular Values

The columns of U and V are orthogonal to one another:

The pseudo-inverse is thus:

Practical Solution of Normal Equations using SVDPractical Solution of Normal Equations using SVD

If we could pre-multiply the forward equation by A-1, the “pseudo-inverse” of A, we could get our answer directly:

The pseudo-inverse is:

where

Practical Solution of Normal Equations using SVDPractical Solution of Normal Equations using SVD

If we could pre-multiply the forward equation by A-1, the “pseudo-inverse” of A, we could get our answer directly:

The pseudo-inverse is:

And the parameter uncertainty covariance

matrix is:

with

Gaussian Probability and Least SquaresGaussian Probability and Least Squares

Residuals vector:

Probability of ri :

Likelihood of r :

N.B.: Only true if residuals are uncorrelated (independent).

PredictionsObservations

Maximum LikelihoodMaximum Likelihood

Log-Likelihood of r :

Goodness-of-fit: 2 for N-M degrees of freedom has a known distribution, so regression models such as this can be judged on the probability of getting a given value of 2.

Probability and Least SquaresProbability and Least Squares

• Why should we expect Gaussian residuals?

Random ProcessesRandom Processesz1 <- runif(5000)

Random ProcessesRandom Processeshist(z1)

Random ProcessesRandom Processesz1 <- runif(5000) z2 <- runif(5000)

What is the distribution of (z1 + z2) ?What is the distribution of (z1 + z2) ?

Triangular DistributionTriangular Distributionhist(z1+z2)

Central Limit TheoremCentral Limit TheoremThere are more ways to get a central value than an extreme one.

Probability and Least SquaresProbability and Least Squares

• Why should we expect Gaussian residuals?

(1) Because the Central Limit Theorem is on our side.

(2) Note that the LS solution is always a minimum variance solution, which is useful by itself. The “maximum-likelihood” interpretation is more of a goal than a reality.

Weighted Least Squares:Weighted Least Squares:More General “Data” ErrorsMore General “Data” Errors

Minimizing the 2 is equivalent to minimizing a cost functioncontaining a covariance matrix C of data errors:

The data error covariance matrix is often taken to be diagonal. This means that you put different levels of confidence on different observations (confidence assigned by assessing both measurement error and amount of trust in your basis functions and linear model). Note that this structure still assumes independence between the residuals.

Covariate Data ErrorsCovariate Data Errors

Recall cost function:

Now allow off-diagonal covariances in C.

N.B. ij = ji and

ii = i2.

Multivariate normalPDF: J propagates

without trouble intothe likelihood expression.

Minimizing J still maximizes the likelihood

Fundamental Trick forFundamental Trick forWeighted and Generalized Least SquaresWeighted and Generalized Least Squares

Transform system (A,b,C) with data covariance matrix C into system (A’,b’,C’), where C’ is the identity matrix:

The Cholesky decomposition computes a “matrix square root” such that if R=chol(C), then C=RR.

You can then solve the Ordinary Least Squares problem A’x = b’, using for instance the SVD method. Note that x remains in regular, untransformed space.