This preview shows pages 1–4. Sign up to view the full content.
1
•
Issues of noise in experimental data: illustration
•
Projection operator
•
Writing the regression equations
•
Getting the pseudo inverse
•
An example: computing pi
Lecture 10: Regression
•
A least square error justification
Example of taxi data
Example of taxi data
Example of taxi data
Example of taxi data
Example of taxi data
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document 2
Example of taxi data
Main problem of noisy data: regression
If I have a set of data points, and I try to ‘fit’ a line through the cloud of
points, what is the best way to fit that line?
Main problem of noisy data: regression
If I have a set of data points, and I try to ‘fit’ a line through the cloud of
points, what is the best way to fit that line?
Main problem of noisy data: regression
If I have a set of data points, and I try to ‘fit’ a line through the cloud of
points, what is the best way to fit that line?
Main problem of noisy data: regression
Main question: “what does good mean”, i.e. what is the criterion for
being a “good” approximation.
•
Avector
Y
of “data”
•
ˆ
Y
projected along the red line
Main problem of noisy data: regression
What does “projected” mean? It means that
ˆ
Y
−
Y
is orthogonoal to
ˆ
Y
This means that from all the points along the red line, the point
is the one which minimizes the distance to
ˆ
Y
Y
3
Main problem of noisy data: regression
What does “projected” mean? It means that
ˆ
Y
−
Y
is orthogonoal to
ˆ
Y
This means that from all the points along the red line, the point
is the one which minimizes the distance to
ˆ
Y
Y
Orthogonality of vectors?
Vectors are orthogonal if their dot product is zero
ˆ
Y
−
Y
is orthogonoal to
ˆ
Y
Means that
dot(
ˆ
Y
−
Y
,
ˆ
Y
)=0
In other words:
ˆ
Y
T
(
ˆ
Y
−
Y
)=0
Where as in the previous chapter
on linear algebra, the superscript T
means transpose
So in summary: to find the projection
In order to find the projection
of a vector
on the red line, one
needs to check the following formula:
ˆ
Y
T
(
ˆ
Y
−
Y
ˆ
Y
Y
Now we can define regression
We consider data points
(
x
i
,y
i
)
y
i
We want to construct an estimator,
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 09/07/2011 for the course ENGIN 7 taught by Professor Horowitz during the Spring '08 term at University of California, Berkeley.
 Spring '08
 HOROWITZ

Click to edit the document details