eece 522 notes_02 ch_1 intro to estimation
TRANSCRIPT
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 1/15
1/15
Ch. 1 Introduction to Estimation
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 2/15
2/15
An Example Estimation Problem: DSB RxS ( f )
f o
M ( f )
f f –f o
)2cos()(),;( oooo t f t m f t s φ π φ +=
Audio
Amp
BPF
&
Amp
)()()( t wt st x +=X
)ˆˆ2cos( oot f φ π +
Est. Algo.
Electronics Adds
Noise w(t )
(usually “white”)
oo f φ ̂&ˆ
f
)(ˆ f M
Oscillator
w/ oo f φ ̂&ˆ
Goal: Given
Find Estimates(that are optimal in some sense)
)(),;()( t w f t st x oo += φ Describe with
Probability Model:
PDF & Correlation
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 3/15
3/15
Discrete-Time Estimation Problem
These days, almost always work with samples of the observedsignal (signal plus noise): ][],;[][ nw f nsn x oo += φ
Our “Thought” Model: Each time you “observe” x[n] itcontains same s[n] but different “realization” of noise w[n],
so the estimate is different each time.oo f φ ̂&ˆ are RVs
Our Job: Given finite data set x[0], x[1], … x[ N -1]
Find estimator functions that map data into estimates:
)(])1[,],1[],0[(ˆ
)(])1[,],1[],0[(ˆ
22
11
x
x
g N x x xg
g N x x xg f
o
o
=−=
=−=
…
…
φ
These are RVs…
Need to describe w/
probability model
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 4/15
4/15
PDF of Estimate
Because estimates are RVs we describe them with a PDF…Will depend on:
1. structure of s[n]
2. probability model of w[n]
3. form of est. function g(x))ˆ( o f p
o
f ̂o f
Mean measures centroid
Std. Dev. & Variancemeasure spread
Desire:
( ) small}ˆ{ˆ
}ˆ{
22ˆ =
−=
=
oo f
oo
f E f E
f f E
o
σ
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 5/15
5/15
1.2 Mathematical Estimation Problem
General Mathematical Statement of Estimation Problem:For… Measured Data x = [ x[0] x[1] … x[ N -1] ]
Unknown Parameter = [θ 1 θ 2 … θ p ]
is Not Random
x is an N -dimensional random data vector
Q: What captures all the statistical information needed for anestimation problem ?
A: Need the N -dimensional PDF of the data, parameterized by
);( θx pIn practice, not given PDF!!!
Choose a suitable model
• Captures Essence of Reality• Leads to Tractable Answer We’ll use p(x; ) to find )(ˆ xθ g=
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 6/15
6/15
Ex. Estimating a DC Level in Zero Mean AWGN
]0[]0[ w x +=θ Consider a single data point is observedGaussian
zero mean
variance σ2
~ N (θ , σ2)
So… the needed parameterized PDF is:
p( x[0];θ ) which is Gaussian with mean of θ
So… in this case the parameterization changes the data PDF mean:
θ 1
p( x[0];θ 1)
x[0] θ 2
p( x[0];θ 2)
x[0] θ 3
p( x[0];θ 3)
x[0]
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 7/15
7/15
Ex. Modeling Data with Linear Trend
See Fig. 1.6 in Text
Looking at the figure we see what looks like a linear trend
perturbed by some noise…
So the engineer proposes signal and noise models:
[ ] ][][],;[
nw Bn An x B Ans
++= "#"$%
Signal Model: Linear Trend Noise Model: AWGN
w/ zero mean
AWGN = “Additive White Gaussian Noise”
“White” = x[n] and x[m] are uncorrelated for n ≠ m } Iwwww 2))(( σ =−− T E
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 8/15
8/15
Typical Assumptions for Noise Model
• W and G is always easiest to analyze
– Usually assumed unless you have reason to believe otherwise
– Whiteness is usually first assumption removed
– Gaussian is less often removed due to the validity of Central Limit Thm
• Zero Mean is a nearly universal assumption
– Most practical cases have zero mean
– But if not… µ += ][][ nwnw zm
Non-Zero Mean of µ Zero Mean Now group into signal model
• Variance of noise doesn’t always have to be known to make an
estimate– BUT, must know to assess expected “goodness” of the estimate
– Usually perform “goodness” analysis as a function of noise variance (or
SNR = Signal-to-Noise Ratio)– Noise variance sets the SNR level of the problem
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 9/15
9/15
Classical vs. Bayesian Estimation Approaches
If we view (parameter to estimate) as Non-Random→ Classical Estimation
Provides no way to include a priori information about
If we view (parameter to estimate) as Random
→ Bayesian Estimation
Allows use of some a priori PDF on
The first part of the course: Classical Methods
• Minimum Variance, Maximum Likelihood, Least Squares
Last part of the course: Bayesian Methods
• MMSE, MAP, Wiener filter, Kalman Filter
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 10/15
10/15
1.3 Assessing Estimator PerformanceCan only do this when the value of is known:
• Theoretical Analysis, Simulations, Field Tests, etc.
is a random variableRecall that the estimate )(ˆ xg
Thus it has a PDF of its own… and that PDF completely displays
the quality of the estimate.
Illustrate with 1-D parameter case
θ θ ̂
)ˆ(θ p
Often just capture quality through mean and variance of )(ˆ xg
Desire:
( ) small}ˆ{ˆ
}ˆ{
22ˆ
ˆ
=
−=
==
θ θ σ
θ θ
θ
θ
E E
E m If this is true:
say estimate is
“unbiased
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 11/15
11/15
Equivalent View of Assessing Performance
)ˆ
(ˆ
ee +=−= θ θ θ θ Define estimation error:
RV RV Not RV
Completely describe estimator quality with error PDF: p(e)
p(e)
e
Desire:
( ){ } small}{
0}{
22 =−=
==
e E e E
e E m
e
e
σ
If this is true:
say estimate is
“unbiased
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 12/15
12/15
Example: DC Level in AWGN
Model: x 1,,1,0],[][ −=+= N nnw An…
Gaussian, zero mean, variance σ2
White (uncorrelated sample-to-sample)
PDF of an individual data sample:
−−=
2
2
2 2
)][(exp
2
1])[(
σ πσ
Ai xi x p
Uncorrelated Gaussian RVs are Independent…so joint PDF is the product of the individual PDFs:
−−=
−−= ∑∏
−
=−
=2
1
0
2
2 / 2
1
02
2
2 2
)][(exp
)2(
1
2
)][(exp
2
1)(
σ πσ σ πσ
N
n
N
N
n
An x An x
p x
( property: prod of exp’s gives sum inside exp )
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 13/15
13/15
Each data sample has the same mean ( A), which is the thing we
are trying to estimate… so, we can imagine trying to estimate A by finding the sample mean of the data:
Statistics ∑−
=
=1
0
][1ˆ
N
n
n x N
A
Prob. Theory
Let’s analyze the quality of this estimator…• Is it unbiased?
A A E
i x E N
n x N
E A E n
A
N
n
=⇒
=
=
∑∑ =
−
=
}ˆ{
]}[{1
][1
}ˆ{1
0#$%
Yes! Unbiased!
N A
N
N
N n x
N n x
N A
N
n
N
n
N
n
2
2
21
0
2
2
1
02
1
0
)ˆvar(
1])[var(
1][
1var)ˆvar(
σ
σ σ
=⇒
===
= ∑∑∑
−
=
−
=
−
=
Can make var small by increasing N!!!
Due to Indep.
(white & Gauss.
⇒ Indep.)
• Can we get a small variance?
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 14/15
14/15
Theoretical Analysis vs. Simulations
• Ideally we’d like to be always be able to theoreticallyanalyze the problem to find the bias and variance of the
estimator
– Theoretical results show how performance depends on the problemspecifications
• But sometimes we make use of simulations
– to verify that our theoretical analysis is correct
– sometimes can’t find theoretical results
7/30/2019 EECE 522 Notes_02 Ch_1 Intro to Estimation
http://slidepdf.com/reader/full/eece-522-notes02-ch1-intro-to-estimation 15/15
15/15
Course Goal = Find “Optimal” Estimators
• There are several different definitions or criteria for optimality!
• Most Logical: Minimum MSE (Mean-Square-Error)
– See Sect. 2.4
– To see this result: ( )
)(}ˆvar{
ˆ)ˆ(
2
2
θ θ
θ θ θ
b
E mse
+=
−=
( )
( ) ( )[ ][ ] { }
)(}ˆvar{
)(}ˆ{ˆ)(}ˆ{ˆ
}ˆ
{}ˆ
{ˆ
ˆ)ˆ(
2
2
0
2
2
2
θ θ
θ θ θ θ θ θ
θ θ θ θ
θ θ θ
b
b E E b E E
E E E
E mse
+=
+−+
−=
−+−=
−=
=
" "#" "$%
θ θ θ −= }ˆ{)( E b
Bias
Although MSE makes sense, estimates usually rely on