how do i know the answer if i’m not sure of the question? putting robustness into estimation k. e....
TRANSCRIPT
How do I know the answer if I’m not sure of the
question?
Putting robustness into estimation
K. E. Schubert11/7/00
Familiar Picture?
Basic Problem
Picture of something that has been blurred
If I know how it was blurred then I should be able to clean it up
If system is invertible then I can get the original
x b
A
xb
A†
Familiar Picture
Encountering Resistance
Consider a simpler problem.– Unknown resistor.– Take current and voltage measurements.– Plot them out.– Want to fit a line to the points.
No measurement is perfect.– No exact fit to all the points.– Want “best” fit.
Measured Values
0 2 4 6 8 10 12 14 16 180
2
4
6
8
10
12
voltage [V]
Unknown Resistor
Gauss’ Stellar Problem
Orbit of Ceres. Errors were in people’s measurements Consider distance from the measurements
to the equation to fit minimize the square of this distance
– min ||Ax-b||2
x=(ATA)
-1A
Tb=A
†b
Understanding Solution
In our problem A, b are vectors Finding nearest scaled A to b Projection
b
A
Ax
Ax-b
Resistor Solved
Want to find slope, 1/R i=(1/R)v Ax=b A vector of voltages b vector of currents x is slope 1/R=v
†i
Best line
0 2 4 6 8 10 12 14 16 180
2
4
6
8
10
12
voltage [V]
Unknown Resistor
Reasonable Question
What if I considered v=iR? Errors assumed in v now! R=i
†v
How do the measured resistances compare?
Comparison of Methods
0 2 4 6 8 10 12 14 16 180
2
4
6
8
10
12
voltage [V]
Unknown Resistor
Errors in Both
A has errors (actual is A+dA) Want to minimize distance
– min ||(A+dA)x-b||2
Need to know something about dA Worst dA in bounded region Best dA in bounded region The dA that makes Ax=b consistent
Worst in a Bounded Region
Keep worst case ok, rest will be fine
||dA||< (bounded region) Projection to farthest A+dA
b
A
(A+dA)x
(A+dA)x-bdA
Best in a Bounded Region
Pick best dA but limit options
||dA||< (bounded region) Projection to nearest A+dA
b
A(A+dA)x
(A+dA)x-bdA
Consistent Equation (TLS)
Called Total Least Squares Projection nearest to A and b in new space No bound on dA, as big as need!
b
A(A+dA)x
General Regression Problems
All of the techniques mentioned so far fall into the general category of regression (including least squares)
Find a solution for most by taking the gradient and setting it equal to zero
x=(ATA+I)
-1A
Tb
Equation for , which is solved by finding the roots of the equation (Newton’s or bisection)
Resistor by TLS
0 2 4 6 8 10 12 14 16 180
2
4
6
8
10
12
voltage [V]
Unknown Resistor
Simple Picture
Consider a city skyline.– Only consider outline of buildings.– Height is a function of horizontal distance.
Nice one dimensional picture.
Hazy Day
Smog and haze blur the image.– Rounds the corners off.– Want to get the corners back.
Actual Measured
0 10 20 30 40 50 60 70 80 90 1000
0.5
1
1.5
2
2.5
3
3.5
Least Squares Fails!
Blurring works like a Gaussian distribution Don’t know the exact blur
Measured Least Squares
0 10 20 30 40 50 60 70 80 90 100-2000
-1500
-1000
-500
0
500
1000
1500
2000
Least Squares Solution
TLS Too Optimistic!
TLS assumes things are consistent Allows dA to be large
MeasuredTLS
0 10 20 30 40 50 60 70 80 90 100-60
-40
-20
0
20
40
60
More Robust Solutions
Picking a solution with some restrictions yields good results.
Actual MeasuredMinMin MinBE
0 10 20 30 40 50 60 70 80 90 100-0.5
0
0.5
1
1.5
2
2.5
3
3.5
Sample
Sig
na
l
Conclusions
Least Squares has nice properties and generally works well.
Problems can arise in simple problems.– Fundamental errors
Must account for errors in basic system. Robust ~ works well for all nearby systems
– Can’t do as well or as bad (compromise)