(51 intermediate revisions by the same user not shown)
Line 2: Line 2:
 
[[Category:ECE662]]
 
[[Category:ECE662]]
  
<center><font size= 4>
+
<center><font size= 5>
[[ECE662_Bayesian_Parameter_Estimation_S14_SF|Bayesian Parameter Estimation]]
+
[[ECE662_Bayesian_Parameter_Estimation_S14_SF|Bayesian Parameter Estimation: Gaussian Case]]
 
</font size>
 
</font size>
  
Line 11: Line 11:
 
</center>
 
</center>
  
[[Media:slecture_ECE662_Whitening_and_Coloring_Transforms_S14_MH.pdf|PDF version]]
 
 
----
 
----
 
----
 
----
Line 27: Line 26:
 
</center>
 
</center>
  
Furthermore, <math>p(x|D)</math> can be computed as:
+
As the above equation suggests, we can use the information provided by the training data to help determine both the class-conditional densities and the priori probabilities.
 +
 
 +
Furthermore, since we are treating supervised case, we can separate the training samples by class into c subsets <math>\mathcal{D}_1, \mathcal{D}_2, ..., \mathcal{D}_c</math>, accordingly:
 +
 
 +
<center>
 +
<math>P(w_i|x,D) = \frac{p(x|w_i,D_i)P(w_i)}{\sum_{j = 1}^c p(x|w_j,D_j)P(w_j)}</math>
 +
</center>
 +
 
 +
Now, assume <math>p(x)</math> has a parameter form. We are given a set of <math>N</math> independent samples <math>\mathcal{D} = \{x_1, x_2, ... , x_N \}</math>. View <math>\theta</math> as a random variable. Consider more specifically in continuous case:
 +
 
 +
<math>p(x|D)</math> can be computed as:
 
<center>
 
<center>
<math>p(x|D) = \int p(x|\theta)p(\theta|D)d\theta</math>
+
<math>p(x|D) = \int p(x, \theta|D)d\theta = \int p(x|\theta)p(\theta|D)d\theta</math>
 
</center>
 
</center>
 
----
 
----
Line 37: Line 46:
  
  
It is important to know that:  
+
We first start with a generalized approach which can be applied to any situation in which the unknown density can be parameterized. The basic assumptions are as follows:
  
1. The form of the density $p(x|\theta)$ is assumed to be known, but the value of the parameter vector <math>\theta</math> is not known exactly.  
+
1. The form of the density <math>p(x|\theta)</math> is assumed to be known, but the value of the parameter vector <math>\theta</math> is not known exactly.  
  
 
2. The initial knowledge about <math>\theta</math> is assumed to be contained in a known a priori density <math>p(\theta)</math>.  
 
2. The initial knowledge about <math>\theta</math> is assumed to be contained in a known a priori density <math>p(\theta)</math>.  
Line 45: Line 54:
 
3. The rest of the knowledge about <math>\theta</math> is contained in a set <math>\mathcal{D}</math> of n samples <math>x_1, x_2, ... , x_n</math> drawn independently according to the unknown probability density <math>p(x)</math>.
 
3. The rest of the knowledge about <math>\theta</math> is contained in a set <math>\mathcal{D}</math> of n samples <math>x_1, x_2, ... , x_n</math> drawn independently according to the unknown probability density <math>p(x)</math>.
  
Accordingly, based on:
+
Accordingly, already know:
  
 
<center><math>p(x|D) = \int p(x|\theta)p(\theta|D)d\theta</math></center>
 
<center><math>p(x|D) = \int p(x|\theta)p(\theta|D)d\theta</math></center>
  
and Bayes Theorem,
+
and By Bayes Theorem,
  
 
<center><math>p(\theta|D) = \frac{p(D|\theta)p(\theta)}{\int p(D|\theta)p(\theta|D)d\theta}</math></center>
 
<center><math>p(\theta|D) = \frac{p(D|\theta)p(\theta)}{\int p(D|\theta)p(\theta|D)d\theta}</math></center>
  
  
Now, since we are attempting to transform the equation to be based on samples <math>x_i</math>, by independent assumption,  
+
Now, since we are attempting to transform the equation to be based on samples <math>x_k</math>, by independent assumption,  
  
<center><math>p(D|\theta) = \prod_{k = 1}^n p(x_i|\theta)</math></center>
+
<center><math>p(D|\theta) = \prod_{k = 1}^n p(x_k|\theta)</math></center>
  
 
Hence, if a sample <math>\mathcal{D}</math> has n samples, we can denote the sample space as:
 
Hence, if a sample <math>\mathcal{D}</math> has n samples, we can denote the sample space as:
+
<math>\mathcal{D}^n = \{x_1, x_2, ... x_n\}</math>.
<center><math>\mathcal{D}^n = \{x_1, x_2, ... x_n\}</math>.</center>
+
  
 +
Combine the sample space definition with the equation above:
  
  
<center><math>p(D^n|\theta) = p(D^{n-1}|\theta)p(x_n|\theta)</math></center>
+
 
 +
                                <center><math> p(D^n|\theta) = p(D^{n-1}|\theta)p(x_n|\theta) </math></center>
  
 
Using this equation, we can transform the Bayesian Parameter Estimation to:
 
Using this equation, we can transform the Bayesian Parameter Estimation to:
Line 72: Line 82:
  
  
Investigation of Estimator's Accuracy of Bayesian Parameter Estimation: Gaussian Case
 
  
The Univariate Case: <math>p(\mu|\mathcal{D})</math>
+
----
 +
== '''Bayesian Parameter Estimation: Gaussian Case''' ==
 +
 
 +
== ''The Univariate Case: <math>p(\mu|\mathcal{D})</math>'' ==
  
Assumptions:  
+
Consider the case where <math>\mu</math> is the only unknown parameter. For simplicity we assume:
  
<center><math>p(x|\mu) \sim N(\mu, \sigma^2)</math></center>
+
                                        <center><math>p(x|\mu) \sim N(\mu, \sigma^2)</math></center> and
  
 
<center><math>p(\mu) \sim N(\mu_0, \sigma_0^2)</math></center>
 
<center><math>p(\mu) \sim N(\mu_0, \sigma_0^2)</math></center>
  
From the previous section, the following expression could be easily obtained:
+
From the previous section, the following expression could be easily obtained using Bayes' formula:
  
 
<center><math>p(\mu|D) = \alpha \prod_{k = 1}^n p(x_k|\mu)p(\mu)</math></center>
 
<center><math>p(\mu|D) = \alpha \prod_{k = 1}^n p(x_k|\mu)p(\mu)</math></center>
Line 94: Line 106:
 
<center><math>p(u) = \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}]</math></center>
 
<center><math>p(u) = \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}]</math></center>
  
Hence, accordingly,
+
The equation has now become:
  
 
<center><math>p(\mu|D) = \alpha \prod_{k = 1}^n \frac{1}{(2\pi\sigma^2)^{1/2}}exp[-\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}] \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}]</math></center>
 
<center><math>p(\mu|D) = \alpha \prod_{k = 1}^n \frac{1}{(2\pi\sigma^2)^{1/2}}exp[-\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}] \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}]</math></center>
Line 100: Line 112:
 
<center><math>p(\mu|D) = \alpha \prod_{k = 1}^n \frac{1}{(2\pi\sigma^2)^{1/2}} \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}  -\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}]</math></center>
 
<center><math>p(\mu|D) = \alpha \prod_{k = 1}^n \frac{1}{(2\pi\sigma^2)^{1/2}} \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}  -\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}]</math></center>
  
Update the scaling factor to <math>\beta</math>,  
+
Update the scaling factor to <math>\alpha'</math> and <math>\alpha''</math> correspondingly,
  
<center><math>p(\mu|D) = \beta exp \sum_{k=1}^n(-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}  -\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2})</math></center>
+
<center><math>p(\mu|D) = \alpha' exp \sum_{k=1}^n(-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}  -\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2})</math></center>
  
<center><math>p(\mu|D) = \gamma exp [-\frac{1}{2}(\frac{n}{\sigma^2} + \frac{1}{\sigma_0^2})\mu^2  -2(\frac{1}{\sigma^2}\sum_{k=1}^nx_k + \frac{\mu_0}{\sigma_0^2})\mu]</math></center>
+
<center><math>p(\mu|D) = \alpha'' exp [-\frac{1}{2}(\frac{n}{\sigma^2} + \frac{1}{\sigma_0^2})\mu^2  -2(\frac{1}{\sigma^2}\sum_{k=1}^nx_k + \frac{\mu_0}{\sigma_0^2})\mu]</math></center>
  
Furthermore, since
+
With the knowledge of Gaussian distribution:
  
 
<center><math>p(u|D) = \frac{1}{(2\pi\sigma_n^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_n}{\sigma_n})^{2}]</math></center>
 
<center><math>p(u|D) = \frac{1}{(2\pi\sigma_n^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_n}{\sigma_n})^{2}]</math></center>
Line 116: Line 128:
 
Where <math>\bar{x_n}</math> is defined as sample means and <math>n</math> is the sample size.
 
Where <math>\bar{x_n}</math> is defined as sample means and <math>n</math> is the sample size.
  
In order to form a Gaussian distribution, the variance <math>\sigma_n^2</math> associated with <math>\mu_n</math> is defined as:
+
In order to form a Gaussian distribution, the variance <math>\sigma_n^2</math> associated with <math>\mu_n</math> could also be obtained correspondingly as:
  
 
<center><math>\sigma_n^2 = \frac{\sigma_0^2 \sigma^2}{n\sigma_0^2 + \sigma^2}</math></center>
 
<center><math>\sigma_n^2 = \frac{\sigma_0^2 \sigma^2}{n\sigma_0^2 + \sigma^2}</math></center>
  
The Univariate Case: <math>p(x|\mathcal{D})</math>
+
 
 +
Observation:
 +
 
 +
With <math>N \to \infty</math>, <center><math>\sigma_D \to 0</math></center>
 +
<math>p(\mu|D)</math> becomes more sharply peaked around <math>\mu_D</math>
 +
 
 +
== ''The Univariate Case: <math>p(x|\mathcal{D})</math>'' ==
 +
 
  
 
Having obtained the posteriori density for the mean <math>u_n</math> of set <math>\mathcal{D}</math>, the remaining of the task is to estimate the "class-conditional" density for <math>p(x|D)</math>.
 
Having obtained the posteriori density for the mean <math>u_n</math> of set <math>\mathcal{D}</math>, the remaining of the task is to estimate the "class-conditional" density for <math>p(x|D)</math>.
  
Based on the text by \textbf{Duda's},
+
Based on the text Duda's chatpter #3.4.2 and Prof. Mimi's notes:
  
  
 
<center><math>p(x|\mathcal{D}) = \int p(x|\mu)p(\mu|\mathcal{D})d\mu</math></center>
 
<center><math>p(x|\mathcal{D}) = \int p(x|\mu)p(\mu|\mathcal{D})d\mu</math></center>
 +
<center><math>p(x|\mathcal{D}) = \int \frac{1}{\sqrt{2 \pi } \sigma} \exp[{-\frac{1}{2} (\frac{x-\mu}{\sigma})^2}]  \frac{1}{\sqrt{2 \pi } \sigma_n} \exp[{-\frac{1}{2} (\frac{\mu-\mu_n}{\sigma_n})^2}] d\mu</math></center>
  
  
Line 142: Line 162:
 
<center><math>p(x|D) \sim N(\mu_n, \sigma^2 + \sigma_n^2)</math></center>
 
<center><math>p(x|D) \sim N(\mu_n, \sigma^2 + \sigma_n^2)</math></center>
  
 
+
----
\subsection{Experiment of Bayesian Parameter Estimation}
+
 
+
\paragraph{Design}
+
 
+
Assume n samples were obtained from the class <math>\mathcal{D}</math> of unknown mean <math>\mu</math> (known <math>\sigma</math>). Assume,
+
 
+
<math>p(x|\mu) \sim N(\mu, \sigma^2)</math>
+
 
+
<math>p(\mu) \sim N(\mu_0, \sigma_0^2)</math>
+
 
+
While <math>\sigma = \sigma_0 = constant</math>, and <math>\mu_0 = 0</math> (It does not matter what <math>\mu_0</math> it was assumed to be, this will be verified shortly after). Based on the sample data <math>x_i \in \mathcal{D}, i = 1,2,3,...,n</math>, <math>\mu</math> is desired to be estimated.
+
 
+
The following results will be obtained:
+
\begin{enumerate}
+
\item The impact of $\mu_0$ on estimated $\hat{\mu}$
+
\item The impact of sample size $n$ have on estimation accuracy
+
\end{enumerate}
+
 
+
\paragraph{Results}
+
\begin{center}
+
\includegraphics[scale=1]{BPE_1.png}
+
 
+
Figure 21. The impact of $\mu_0$ on estimated $\hat{\mu}$ averaged over 50 samples
+
 
+
\includegraphics[scale=1]{BPE_2.png}
+
 
+
Figure 22. The impact of $\mu_0$ on the variance of estimated $\hat{\mu}$ over 50 samples
+
 
+
 
+
\end{center}
+
 
+
The estimated mean is shifting up with $\mu_0$ increasing. \textbf{Based on the experiment it can be concluded that the most 'accurate' estimate could be obtained if <math>\mu_0 = \mu</math>. But, according to the plot, even if the $\mu_0$ is different different from $\mu$, the error of estimation is still acceptable. (In our case, within [-0.1,+0.06] region)} However, the variance of estimated mean could be assumed to be identical as the \textbf{real empirical mean}.
+
 
+
 
+
\begin{center}
+
\includegraphics[scale=0.7]{e23456.png}
+
 
+
Figure 23. The impact of sample size $n$ have on estimation shape accuracy (sample sizes = 2,3,4,5,6)
+
 
+
 
+
\includegraphics[scale=0.6]{ece662_14.png}
+
 
+
Figure 24. The impact of sample size $n$ have on estimation shape accuracy (sample sizes = 4,10,20,50,100)
+
 
+
 
+
\paragraph{Conclusion} Figure 23. and Figure 24. have demonstrated that with insufficient sample size the result would be really poor regarding prediction of points distribution. 
+
 
+
<center>[[Image:fig3_summ_mh.png|680px|thumb|left|Fig 3: Summary diagram of whitening and coloring process.]]</center>
+
 
+
 
+
 
----
 
----
  

Latest revision as of 07:31, 29 April 2014


Bayesian Parameter Estimation: Gaussian Case

A slecture by ECE student Shaobo Fang

Loosely based on the ECE662 Spring 2014 lecture material of Prof. Mireille Boutin.



Introduction: Bayesian Estimation

According to Chapter #3.3 (Duda's book), although the answers we get by BPE will generally be nearly identical to those obtained by maximum likelihood estimation, the conceptual difference is significant. For maximum likelihood estimation, the parameter $ \theta $ is a fixed while in Bayersian estimation $ \theta $ is considered to be a random variable.

By definition, given samples class $ \mathcal{D} $, Bayes' formula then becomes:


$ P(w_i|x,D) = \frac{p(x|w_i,D)P(w_i|D)}{\sum_{j = 1}^c p(x|w_j,D)P(w_j|D)} $

As the above equation suggests, we can use the information provided by the training data to help determine both the class-conditional densities and the priori probabilities.

Furthermore, since we are treating supervised case, we can separate the training samples by class into c subsets $ \mathcal{D}_1, \mathcal{D}_2, ..., \mathcal{D}_c $, accordingly:

$ P(w_i|x,D) = \frac{p(x|w_i,D_i)P(w_i)}{\sum_{j = 1}^c p(x|w_j,D_j)P(w_j)} $

Now, assume $ p(x) $ has a parameter form. We are given a set of $ N $ independent samples $ \mathcal{D} = \{x_1, x_2, ... , x_N \} $. View $ \theta $ as a random variable. Consider more specifically in continuous case:

$ p(x|D) $ can be computed as:

$ p(x|D) = \int p(x, \theta|D)d\theta = \int p(x|\theta)p(\theta|D)d\theta $


Bayesian Parameter Estimation: General Theory

We first start with a generalized approach which can be applied to any situation in which the unknown density can be parameterized. The basic assumptions are as follows:

1. The form of the density $ p(x|\theta) $ is assumed to be known, but the value of the parameter vector $ \theta $ is not known exactly.

2. The initial knowledge about $ \theta $ is assumed to be contained in a known a priori density $ p(\theta) $.

3. The rest of the knowledge about $ \theta $ is contained in a set $ \mathcal{D} $ of n samples $ x_1, x_2, ... , x_n $ drawn independently according to the unknown probability density $ p(x) $.

Accordingly, already know:

$ p(x|D) = \int p(x|\theta)p(\theta|D)d\theta $

and By Bayes Theorem,

$ p(\theta|D) = \frac{p(D|\theta)p(\theta)}{\int p(D|\theta)p(\theta|D)d\theta} $


Now, since we are attempting to transform the equation to be based on samples $ x_k $, by independent assumption,

$ p(D|\theta) = \prod_{k = 1}^n p(x_k|\theta) $

Hence, if a sample $ \mathcal{D} $ has n samples, we can denote the sample space as: $ \mathcal{D}^n = \{x_1, x_2, ... x_n\} $.

Combine the sample space definition with the equation above:


$ p(D^n|\theta) = p(D^{n-1}|\theta)p(x_n|\theta) $

Using this equation, we can transform the Bayesian Parameter Estimation to:

$ p(\theta|D^n) = \frac{p(x_n|\theta)p(\theta|D^{n-1})}{\int p(x_n|\theta)p(\theta|D^{n-1})d\theta} $




Bayesian Parameter Estimation: Gaussian Case

The Univariate Case: $ p(\mu|\mathcal{D}) $

Consider the case where $ \mu $ is the only unknown parameter. For simplicity we assume:

$ p(x|\mu) \sim N(\mu, \sigma^2) $
and
$ p(\mu) \sim N(\mu_0, \sigma_0^2) $

From the previous section, the following expression could be easily obtained using Bayes' formula:

$ p(\mu|D) = \alpha \prod_{k = 1}^n p(x_k|\mu)p(\mu) $

Where $ \alpha $ is a factorization factor independent of $ \mu $.

Now, substitute $ p(x_k|\mu) $ and $ p(u) $ with:

$ p(x_k|\mu) = \frac{1}{(2\pi\sigma^2)^{1/2}}exp[-\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}] $
$ p(u) = \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}] $

The equation has now become:

$ p(\mu|D) = \alpha \prod_{k = 1}^n \frac{1}{(2\pi\sigma^2)^{1/2}}exp[-\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}] \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2}] $
$ p(\mu|D) = \alpha \prod_{k = 1}^n \frac{1}{(2\pi\sigma^2)^{1/2}} \frac{1}{(2\pi\sigma_0^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2} -\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}] $

Update the scaling factor to $ \alpha' $ and $ \alpha'' $ correspondingly,

$ p(\mu|D) = \alpha' exp \sum_{k=1}^n(-\frac{1}{2}(\frac{\mu-\mu_0}{\sigma_0})^{2} -\frac{1}{2}(\frac{x_k-\mu}{\sigma})^{2}) $
$ p(\mu|D) = \alpha'' exp [-\frac{1}{2}(\frac{n}{\sigma^2} + \frac{1}{\sigma_0^2})\mu^2 -2(\frac{1}{\sigma^2}\sum_{k=1}^nx_k + \frac{\mu_0}{\sigma_0^2})\mu] $

With the knowledge of Gaussian distribution:

$ p(u|D) = \frac{1}{(2\pi\sigma_n^2)^{1/2}}exp[-\frac{1}{2}(\frac{\mu-\mu_n}{\sigma_n})^{2}] $

Finally, the estimate of $ u_n $ can be obtained:

$ \mu_n = (\frac{n\sigma_0^2}{n\sigma_0^2 + \sigma^2})\bar{x_n} + \frac{\sigma^2}{n\sigma_0^2 + \sigma^2}\mu_0 $

Where $ \bar{x_n} $ is defined as sample means and $ n $ is the sample size.

In order to form a Gaussian distribution, the variance $ \sigma_n^2 $ associated with $ \mu_n $ could also be obtained correspondingly as:

$ \sigma_n^2 = \frac{\sigma_0^2 \sigma^2}{n\sigma_0^2 + \sigma^2} $


Observation:

With $ N \to \infty $,
$ \sigma_D \to 0 $

$ p(\mu|D) $ becomes more sharply peaked around $ \mu_D $

The Univariate Case: $ p(x|\mathcal{D}) $

Having obtained the posteriori density for the mean $ u_n $ of set $ \mathcal{D} $, the remaining of the task is to estimate the "class-conditional" density for $ p(x|D) $.

Based on the text Duda's chatpter #3.4.2 and Prof. Mimi's notes:


$ p(x|\mathcal{D}) = \int p(x|\mu)p(\mu|\mathcal{D})d\mu $
$ p(x|\mathcal{D}) = \int \frac{1}{\sqrt{2 \pi } \sigma} \exp[{-\frac{1}{2} (\frac{x-\mu}{\sigma})^2}] \frac{1}{\sqrt{2 \pi } \sigma_n} \exp[{-\frac{1}{2} (\frac{\mu-\mu_n}{\sigma_n})^2}] d\mu $


$ p(x|\mathcal{D}) = \frac{1}{2\pi\sigma\sigma_n} exp [-\frac{1}{2} \frac{(x-\mu)}{\sigma^2 + \sigma_n^2}]f(\sigma,\sigma_n) $


Where $ f(\sigma, \sigma_n) $ is defined as:


$ f(\sigma,\sigma_n) = \int exp[-\frac{1}{2}\frac{\sigma^2 + \sigma_n^2}{\sigma^2 \sigma_n ^2}(\mu - \frac{\sigma_n^2 x+\sigma^2 \mu_n}{\sigma^2+\sigma_n^2})^2]d\mu $

Hence, $ p(x|D) $ is normally distributed as:

$ p(x|D) \sim N(\mu_n, \sigma^2 + \sigma_n^2) $


References

[1]. Mireille Boutin, "ECE662: Statistical Pattern Recognition and Decision Making Processes," Purdue University, Spring 2014.

[2]. R. Duda, P. Hart, Pattern Classification. Wiley-Interscience. Second Edition, 2000.

Alumni Liaison

Correspondence Chess Grandmaster and Purdue Alumni

Prof. Dan Fleetwood