Difference between revisions of "Aufgaben:Exercise 3.5Z: Kullback-Leibler Distance again"

From LNTwww
 
(20 intermediate revisions by 4 users not shown)
Line 1: Line 1:
  
{{quiz-Header|Buchseite=Informationstheorie/Einige Vorbemerkungen zu zweidimensionalen Zufallsgrößen
+
{{quiz-Header|Buchseite=Information_Theory/Some_Preliminary_Remarks_on_Two-Dimensional_Random_Variables
 
}}
 
}}
  
[[File:P_ID2762__Inf_Z_3_4.png|right|]]
+
[[File:P_ID2762__Inf_Z_3_4.png|right|frame|Determined probability mass functions]]
Die Wahrscheinlichkeitsfunktion lautet:
+
The probability mass function is:
 +
:$$P_X(X) = \big[\hspace{0.03cm}0.25\hspace{0.03cm}, \hspace{0.15cm} 0.25\hspace{0.15cm},\hspace{0.15cm} 0.25 \hspace{0.03cm}, \hspace{0.15cm} 0.25\hspace{0.03cm}\big]\hspace{0.05cm}.$$
 +
The random variable  $X$  is thus characterised by
 +
* the symbol set size  $M=4$,
 +
* equal probabilities  $P_X(1) = P_X(2) = P_X(3) = P_X(4) = 1/4$ .
  
$$P_Y(X) = [\hspace{0.03cm}0.25\hspace{0.03cm}, \hspace{0.03cm} 0.25\hspace{0.03cm},\hspace{0.03cm} 0.25 \hspace{0.03cm}, \hspace{0.03cm} 0.25\hspace{0.03cm}]\hspace{0.05cm}$$
 
Die Zufallsgröße $X$ ist also gekennzeichnet
 
  
:* durch den Symbolumfang $M=4$,
+
The random variable  $Y$  is always an approximation for  $X$:
:* mit gleichen Wahrscheinlichkeiten.
+
*It was obtained by simulation from a uniform distribution, whereby only  $N$  random numbers were evaluated in each case.  This means:  
 +
*$P_Y(1)$, ... , $P_Y(4)$  are not probabilities in the conventional sense.  Rather, they describe  [[Theory_of_Stochastic_Signals/Wahrscheinlichkeit_und_relative_H%C3%A4ufigkeit#Bernoullisches_Gesetz_der_gro.C3.9Fen_Zahlen| relative frequencies]].
  
Die Zufallsgröße $Y$ ist stets eine Näherung für $X$. Sie wurde per Simulation aus einer Gleichverteilung gewonnen, wobei jeweils nur $N$ Zufallswerte ausgewertet wurden. Das heißt: 
 
$P_Y(1)$,...,$P_Y(4)$ sind im herkömmlichen Sinn keine Wahrscheinlichkeiten. Sie beschreiben vielmehr [http://en.lntwww.de/Stochastische_Signaltheorie/Wahrscheinlichkeit_und_relative_H%C3%A4ufigkeit#Bernoullisches_Gesetz_der_gro.C3.9Fen_Zahlen relative Häufigkeiten].
 
  
Das Ergebnis der sechsten Versuchsreihe (mit $N=1000$) ird demnach durch die folgende Wahrscheinlichkeitsfunktion zusammengefasst:
+
The result of the sixth test series  (with  $N=1000)$  is thus summarised by the following probability function:
  
$$P_Y(X) = [\hspace{0.05cm}0.225\hspace{0.05cm}, \hspace{0.05cm} 0.253\hspace{0.05cm},\hspace{0.05cm} 0.250 \hspace{0.05cm}, \hspace{0.05cm} 0.272\hspace{0.05cm}]
+
:$$P_Y(X) = \big [\hspace{0.05cm}0.225\hspace{0.15cm}, \hspace{0.05cm} 0.253\hspace{0.05cm},\hspace{0.15cm} 0.250 \hspace{0.05cm}, \hspace{0.15cm} 0.272\hspace{0.05cm}\big]
\hspace{0.05cm}$$
+
\hspace{0.05cm}.$$
Bei dieser Schreibweise ist bereits berücksichtigt, dass die Zufallsgrößen $X$ und $Y$ auf dem gleichen Alphabet $X =$ {1, 2, 3, 4} basieren.
+
This notation already takes into account that the random variables  $X$  and  $Y$  are based on the same alphabet  $X = \{1,\ 2,\ 3,\ 4\}$.
  
Mit diesen Voraussetzungen gilt für die relative Entropie (englisch: Informational Divergence) zwischen den Wahrscheinlichkeitsfunktionen $P_X(.)$ und $P_Y(.)$ :
+
With these preconditions, the  '''informational divergence'''  between the two probability functions  $P_X(.)$  and  $P_Y(.)$ :
  
$D( P_X || P_Y) = E_X [ log_2 \frac{P_X(X)}{P_Y(Y)}] = \sum\limits_{\mu=1}^M P_X(\mu) . log_2 \frac{P_X(\mu)}{P_Y(\mu)}$
+
:$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = {\rm E}_X \hspace{-0.1cm}\left [ {\rm log}_2 \hspace{0.1cm} \frac{P_X(X)}{P_Y(X)}\right ] \hspace{0.2cm}=\hspace{0.2cm} \sum_{\mu = 1}^{MP_X(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_X(\mu)}{P_Y(\mu)} \hspace{0.05cm}.$$
  
Man bezeichnet $D( P_X || P_Y)$  als Kullback–Leibler–Distanz. Diese ist ein Maß für die Ähnlichkeit zwischen den beiden Wahrscheinlichkeitsfunktionen
+
One calls  $D( P_X\hspace{0.05cm} || \hspace{0.05cm}P_Y)$  the (first)  '''Kullback-Leibler distance'''.
$P_X(.)$ und $P_Y(.)$.  Die Erwartungswertbildung geschieht hier hinsichtlich der (tatsächlich gleichverteilten) Zufallsgröße $X$. Dies wird durch die Nomenklatur $E_X[.]$ angedeutet.
+
*This is a measure of the similarity between the two probability mass functions  $P_X(.)$  and  $P_Y(.)$.   
 +
*The expected value formation occurs here with regard to the  (actually equally distributed)  random variable  $X$.  This is indicated by the nomenclature  ${\rm E}_X\big[.\big]$.
  
Eine zweite Form der Kullback–Leibler–Distanz ergibt sich durch die Erwartungswertbildung  hinsichtlich der Zufallsgröße $Y \Rightarrow E_Y[.]$:
 
  
$D( P_Y || P_X) = E_Y [ log_2 \frac{P_Y(Y)}{P_Y(Y)}] = \sum\limits_{\mu=1}^M P_Y(\mu) . log_2 \frac{P_Y(\mu)}{P_X(\mu)}$
+
A second form of Kullback-Leibler distance results from the formation of expected values with respect to the random variable  $Y$   ⇒    ${\rm E}_Y\big [.\big ]$:
  
'''Hinweis:''' Die Aufgabe bezieht sich auf das [http://en.lntwww.de/Informationstheorie/Einige_Vorbemerkungen_zu_zweidimensionalen_Zufallsgr%C3%B6%C3%9Fen Kapitel 3.1 ] dieses Buches. Die Angaben der Entropie $H(Y)$ und der Kullback–Leibler–Distanz  $D( P_X || P_Y)$  in obiger Grafik sind in „bit” zu verstehen. die mit „???"  versehenen Felder sollen von Ihnen in dieser Aufgabe ergänzt werden.
+
:$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) =  {\rm E}_Y \hspace{-0.1cm} \left [ {\rm log}_2 \hspace{0.1cm} \frac{P_Y(X)}{P_X(X)}\right ] \hspace{0.2cm}=\hspace{0.2cm} \sum_{\mu = 1}^M P_Y(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_Y(\mu)}{P_X(\mu)} \hspace{0.05cm}.$$
  
  
===Fragebogen===
+
 
 +
 
 +
 
 +
Hints:
 +
*The exercise belongs to the chapter  [[Information_Theory/Some_Preliminary_Remarks_on_Two-Dimensional_Random_Variables|Some preliminary remarks on two-dimensional random variables]].
 +
*In particular, reference is made to the page  [[Information_Theory/Some_Preliminary_Remarks_on_Two-Dimensional_Random_Variables#Informational_divergence_-_Kullback-Leibler_distance|Relative entropy – Kullback-Leibler distance]].
 +
*The entropy   $H(Y)$  and the Kullback-Leibler distance   $D( P_X \hspace{0.05cm}|| \hspace{0.05cm}P_Y)$   in the above graph are to be understood in  "bit".
 +
* The fields marked with  "???"  in the graph are to be completed by you in this task.
 +
 +
 
 +
 
 +
===Questions===
  
 
<quiz display=simple>
 
<quiz display=simple>
  
{Welche Entropie besitzt die Zufallsgröße $X$ ?
+
{What is the entropy of the random variable&nbsp; $X$ ?
 
|type="{}"}
 
|type="{}"}
$H(X)$ = { 2 3% } $bit$
+
$H(X)\ = \ $ { 2 1% } $\ \rm bit$
  
{Wie groß sind die Entropien der Zufallsgrößen $Y$ (Näherungen für $X$)?  
+
{What are the entropies of the random variables&nbsp; $Y$&nbsp; $($approximations for&nbsp; $X)$?  
 
|type="{}"}
 
|type="{}"}
$N=1000$ : $H(Y)$ = { 1.9968 1% } $bit$
+
$N=10^3\text{:} \hspace{0.5cm} H(Y) \ = \ $ { 1.9968 1% } $\ \rm bit$
$N=100$ : $H(Y)$ = { 1.941 1% } $bit$
+
$N=10^2\text{:} \hspace{0.5cm} H(Y) \ = \ $ { 1.941 1% } $\ \rm bit$
$N=10$ : $H(Y)$ = { 1.6855 1%  } $bit$
+
$N=10^1\text{:} \hspace{0.5cm} H(Y) \ = \ $ { 1.6855 1%  } $\ \rm bit$
  
{Berechnen Sie die folgenden Kullback–Leibler–Distanzen.
+
{Calculate the following Kullback-Leibler distances.
 
|type="{}"}
 
|type="{}"}
$N=1000$ : $D( P_X || P_Y)$ = { 3.28 1% } . 10 ( { -3 } )$bit$
+
$N=10^3\text{:} \hspace{0.5cm} D( P_X \hspace{0.05cm}|| \hspace{0.05cm}  P_Y) \ = \ $ { 0.00328 1% } $\ \rm bit$
$N=100$ : $D( P_X || P_Y)$= { 4.42 1% } . 10 ( { -2 } )$bit$
+
$N=10^2\text{:} \hspace{0.5cm} D( P_X \hspace{0.05cm}|| \hspace{0.05cm}  P_Y) \ = \ $  { 0.0442 1% } $\ \rm bit$
$N=10$ : $D( P_X || P_Y)$= { 3.45 1% } . 10 ( { -1 } )$bit$
+
$N=10^1\text{:} \hspace{0.5cm} D( P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) \ = \ $  { 0.345 1% } $\ \rm bit$
  
{Liefert $D(P_Y||P_X)$ jeweils exakt das gleiche Ergebnis?
+
{Does&nbsp; $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$&nbsp; give exactly the same result in each case?
|type="[]"}
+
|type="()"}
+ Falsch
+
- Yes.
- Richtig
+
+ No.
  
{Welche Aussagen gelten für die Kullback–Leibler–Distanzen bei $N = 4$?
+
{Which statements are true for the Kullback-Leibler distances with&nbsp; $N = 4$?
 
|type="[]"}
 
|type="[]"}
- Es gilt $D(P_X||P_Y) = 0$.
+
- &nbsp; $D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0$&nbsp; is true.
- Es gilt $D(P_X||P_Y) = 0.5 bit$
+
- &nbsp; $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0.5 \ \rm  bit$&nbsp; is true.
+ $D(P_X||P_Y)$ ist unendlich groß
+
+ &nbsp; $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$&nbsp; is infinitely large.
- Es gilt $D(P_Y||P_X) = 0$.
+
- &nbsp; $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0$&nbsp; holds.
+ Es gilt $D(P_Y||P_X) = 0.5 bit$.  
+
+ &nbsp; $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0.5 \ \rm bit$&nbsp; holds.  
- $D(P_Y||P_X)$ ist unendlich groß.  
+
- &nbsp; $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$&nbsp; is infinitely large.
  
{Ändern sich $H(Y)$ und $D(P_X||P_Y)$monoton mit $N$?
+
{Do both&nbsp; $H(Y)$&nbsp; and&nbsp;  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$&nbsp; change monotonically with&nbsp; $N$?
|type="[]"}
+
|type="()"}
+Falsch
+
- Yes,
-Richtig
+
+ No.
  
  
 
</quiz>
 
</quiz>
  
===Musterlösung===
+
===Solution===
 
{{ML-Kopf}}
 
{{ML-Kopf}}
  
'''1.'''Bei gleichen Wahrscheinlichkeiten gilt mit $M = 4$ :
+
'''(1)'''&nbsp; With equal probabilities, and with&nbsp; $M = 4$:  
 
+
:$$H(X) = {\rm log}_2 \hspace{0.1cm} M
$H(X) = log_2  M = 2 (bit)$
+
\hspace{0.15cm} \underline {= 2\,{\rm (bit)}} \hspace{0.05cm}.$$  
 
 
'''2.''' Die Wahrscheinlichkeiten für die empirisch ermittelten Zufallsgrößen $Y$ weichen im Allgemeinen (nicht immer!) von der Gleichverteilung um so mehr ab, je kleiner der Parameter $N$ ist. Man erhält
 
:* $N = 1000 \Rightarrow  P_Y(Y) =  [0.225, 0.253, 0.250, 0.272]$ :
 
 
 
$H(Y) = 0.225 . log_2 \frac{1}{0.225} +0.253. log_2 \frac{1}{0.253} + 0.250 . log_2 \frac{1}{0.250}+ 0.272 . log_2 \frac{1}{0.272} = 1.9968 (bit)$
 
:* $N = 100\Rightarrow  P_Y(Y) = [0.24, 0.16, 0.30, 0.30]$ :
 
 
 
$H(Y) =$......$= 1.9410$
 
:* $N = 10 \Rightarrow  P_Y(Y) = [0.5, 0.1, 0.3, 0.1]$:
 
 
 
$H(Y) =$......$= 1.6855$
 
  
'''3.'''  Die Gleichung für die gesuchte Kullback–Leibler–Distanz lautet:
 
  
$$D(P_X||P_Y) = \sum\limits_{\mu=1}^4 P_X(\mu) . log_2 \frac{P_X(\mu)}{P_Y(\mu)} =$$
 
  
$$= \frac{1/4}{lg(2)} .[lg \frac{0.25}{P_Y(1)}+\frac{0.25}{P_Y(2)}+\frac{0.25}{P_Y(3)} + \frac{0.25}{P_Y(4)}] =$$  
+
'''(2)'''&nbsp; The probabilities for the empirically determined random variables&nbsp; $Y$&nbsp; generally&nbsp; (not always!)&nbsp; deviate from the uniform distribution the more the parameter&nbsp; $N$&nbsp; is smaller.&nbsp; One obtains for
 +
* $N = 1000 \ \ \Rightarrow \ \ P_Y(Y) =  \big [0.225, \ 0.253, \ 0.250, \ 0.272 \big ]$:
 +
:$$H(Y) =  
 +
0.225 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.225} +
 +
0.253 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.253} +
 +
0.250 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.250} +
 +
0.272 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.272}  
 +
\hspace{0.15cm} \underline {= 1.9968\ {\rm (bit)}}  \hspace{0.05cm},$$
 +
* $N = 100 \ \ \Rightarrow \ \  P_Y(Y) = \big[0.24, \ 0.16, \ 0.30,  \ 0.30\big]$:
 +
:$$H(Y) = \hspace{0.05cm}\text{...} \hspace{0.15cm} \underline {= 1.9410\ {\rm (bit)}}  \hspace{0.05cm},$$
 +
* $N = 10 \ \ \Rightarrow \ \  P_Y(Y) =  \big[0.5, \ 0.1, \ 0.3, \ 0.1 \big]$:
 +
:$$H(Y) = \hspace{0.05cm}\text{...} \hspace{0.15cm} \underline {= 1.6855\ {\rm (bit)}}  \hspace{0.05cm}.$$
  
$$=\frac{1}{4 . lg(2)} . [lg \frac{0.25^4}{P_Y(1) . P_Y(2) . P_Y(3) . P_Y(4)}]$$
 
Der Logarithmus zur Basis 2  $\Rightarrow log_2(.)$ wurde zur einfachen Nutzung des Taschenrechners durch den Zehnerlogarithmus  $\Rightarrow lg(.)$  ersetzt. Man erhält die folgenden numerischen Ergebnisse:
 
:* $N=1000$ :
 
  
$$D(P_X||P_Y)=\frac{1}{4 . lg(2)} . [lg \frac{0.25^4}{0,225 . 0,253 . 0,250 . 0,272}] = 3,28 . 10^{-3} (bit)$$
 
:* $N=100$ :
 
  
$$D(P_X||P_Y)=\frac{1}{4 . lg(2)} . [lg \frac{0.25^4}{0,24 . 0,16 . 0,30 . 0,30}] = 4,42 . 10^{-2} (bit)$$
+
'''(3)'''&nbsp; The equation for the Kullback-Leibler distance we are looking for is:
::* $N=100$ :
 
  
$$D(P_X||P_Y)=\frac{1}{4 . lg(2)} . [lg \frac{0.25^4}{0,5 . 0,1. 0,3 . 0,1}] = 3,45. 10^{-1} (bit)$$
+
:$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \sum_{\mu = 1}^{4}  P_X(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_X(\mu)}{P_Y(\mu)}
 +
=  \frac{1/4}{{\rm lg} \hspace{0.1cm}(2)} \cdot
 +
\left [ {\rm lg} \hspace{0.1cm} \frac{0.25}{P_Y(1)} + \frac{0.25}{P_Y(2)} + \frac{0.25}{P_Y(3)} + \frac{0.25}{P_Y(4)}
 +
\right ] $$
 +
:$$\Rightarrow \hspace{0.3cm} D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y)  =  \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot
 +
\left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{P_Y(1) \cdot P_Y(2)\cdot P_Y(3)\cdot P_Y(4)}
 +
\right ] \hspace{0.05cm}.$$
  
 +
The logarithm to the base&nbsp; $ 2$&nbsp; &rArr;  &nbsp; $\log_2(.)$&nbsp; was replaced by the logarithm to the base&nbsp; $ 10$ &nbsp; &rArr;  &nbsp; $\lg(.)$ for easy use of the calculator.
  
'''5.''' Richtig ist Nein, wie am Beispiel $N = 100$ gezeigt werden soll:
+
The following numerical results are obtained:
 +
* for $N=1000$:
 +
:$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot
 +
\left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{0.225 \cdot 0.253\cdot 0.250\cdot 0.272}
 +
\right ] \hspace{0.15cm} \underline {= 0.00328 \,{\rm (bit)}} \hspace{0.05cm},$$
 +
* for $N=100$:  
 +
:$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot
 +
\left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{0.24 \cdot 0.16\cdot 0.30\cdot 0.30}
 +
\right ] \hspace{0.15cm} \underline {= 0.0442 \,{\rm (bit)}}  \hspace{0.05cm},$$
 +
* for $N=10$:
 +
:$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot
 +
\left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{0.5 \cdot 0.1\cdot 0.3\cdot 0.1}
 +
\right ] \hspace{0.15cm} \underline {= 0.345 \,{\rm (bit)}}  \hspace{0.05cm}.$$
  
$$D(P_X||P_Y) = \sum\limits_{\mu=1}^M P_X(\mu) . log_2 \frac{P_X(\mu)}{P_Y(\mu)} =$$
 
  
$$ = 0.24 . log_2 \frac{0.24}{0.25} +0.16. log_2 \frac{16}{0.25} +2 .  0,30  . log_2 \frac{0.30}{0.25} = 0.0407 (bit)$$
 
In der Teilaufgabe (c) haben wir stattdessen $D(P_X||P_Y)$ = 0.0442 erhalten. Das bedeutet auch: Der Name „Distanz” ist etwas irreführend. Danach würde man eigentlich $D(P_Y||P_X)$ = $D(P_X||P_Y)$ erwarten.
 
  
'''6.'''  Mit $P_Y(X) = [0, 0.25, 0.5, 0.25]$ erhält man:
+
'''(4)'''&nbsp; Correct is&nbsp; <u>'''No'''</u>,&nbsp; as will be shown by the example&nbsp; $N = 100$&nbsp;:
 +
:$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) =  \sum_{\mu = 1}^M P_Y(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_Y(\mu)}{P_X(\mu)} = 0.24\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.24}{0.25} + 0.16\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.16}{0.25} +2 \cdot 0.30\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.30}{0.25}  = 0.0407\ {\rm (bit)}\hspace{0.05cm}.$$
  
$$D(P_X||P_Y) = 0,25 . log_2 \frac{0.25}{0} +2 . 0,25 . log_2 \frac{0.25}{0.25} +0,25 . log_2 \frac{0.25}{0.50}$$
+
*In subtask&nbsp; '''(3)'''&nbsp; we got&nbsp; $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0.0442$&nbsp; instead.  
 +
*This also means: &nbsp; The designation „distance” is somewhat misleading.  
 +
*According to this, one would actually expect&nbsp; $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X) = D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$&nbsp;.
  
Aufgrund des ersten Terms ergibt sich für $D(P_X||P_Y)$ ein unendlich großer Wert. Für die zweite Kullback–Leibler–Distanz gilt:
 
  
$$D(P_Y||P_X) = 0 . log_2 \frac{0}{0.25} +2 . 0,25 . log_2 \frac{0.25}{0.25} +0,25 . log_2 \frac{0.5}{0.25}$$
 
  
Nach einer Grenzwertbetrachtung erkennt man, dass der erste Term das Ergebnis $0$ liefert. Auch der zweite Term ergibt sich zu $0$, und man erhält als Endergebnis:
+
[[File:P_ID2763__Inf_Z_3_4e.png|right|frame|Probability function, entropy and Kullback-Leibler distance]]
 +
'''(5)'''&nbsp; With&nbsp; $P_Y(X) = \big [0, \ 0.25, \ 0.5, \ 0.25 \big ]$&nbsp; one obtains:
 +
:$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0} + 2 \cdot 0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0.25}+0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0.50}\hspace{0.05cm}.$$
  
$D(P_Y||P_X) = 0,50 . log_2(2) = 0.5 (bit)$
+
*Because of the first term, the value of&nbsp; $D(P_X\hspace{0.05cm}|| \hspace{0.05cm}P_Y)$&nbsp; is infinitely large.
 +
*For the second Kullback-Leibler distance holds:
 +
:$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0\cdot {\rm log}_2 \hspace{0.1cm} \frac{0}{0.25} + 2 \cdot 0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0.25}+
 +
0.50\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.5}{0.25}
 +
\hspace{0.05cm}.$$
  
Richtig sind somit die Aussagen $3$ und $5$. Auch aus diesem Extrembeispiel wird deutlich, dass sich $D(P_Y||P_X)$ stets von $D(P_X||P_Y)$ unterscheidet. Nur für den Sonderfall P_Y = P_X sind beide Kullback–Leibler–Distanzen gleich, nämlich $0$. Die folgende Tabelle zeigt das vollständige Ergebnis dieser Aufgabe.
+
*After looking at the limits, one can see that the first term yields the result&nbsp; $0$&nbsp;.&nbsp; The second term also yields zero, and one obtains as the final result:
 +
:$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0.50\cdot {\rm log}_2 \hspace{0.1cm} (2) \hspace{0.15cm} \underline {= 0.5\,{\rm (bit)}} \hspace{0.05cm}.$$
  
 +
&nbsp; <u>Statements 3 and 5</u> are therefore correct:
 +
*From this extreme example it is clear that&nbsp; $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$&nbsp; is always different from&nbsp; $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$&nbsp;.
 +
*Only for the special case&nbsp; $P_Y \equiv P_X$&nbsp; are both Kullback-Leibler distances equal, namely zero.
 +
*The adjacent table shows the complete result of this task.
  
[[File:P_ID2763__Inf_Z_3_4e.png|right|]]
 
'''7.''' Die richtige Antwort ist Nein. Die Tendenz ist zwar eindeutig: Je größer $N$ ist,
 
  
:* desto mehr nähert sich $H(Y)$ im Prinzip dem Endwert $H(X) = 2$ bit an.
 
  
:* um so kleiner werden die Distanzen $D(P_X||P_Y)$ und $D(P_Y||P_X)$.
 
Man erkennt aus obiger Tabelle aber auch, dass es Ausnahmen gibt:
 
  
:* Die Entropie $H(Y)$ ist für $N = 1000$ kleiner als für $N = 400$,
+
'''(6)'''&nbsp; Correct is again&nbsp; <u>'''No'''</u>. &nbsp; Although the tendency is clear: &nbsp; The larger&nbsp; $N$&nbsp; is,
 +
* the more&nbsp; $H(Y)$&nbsp; approaches in principle the final value&nbsp; $H(X) = 2 \ \rm bit$,
 +
* the smaller the distances&nbsp; $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$&nbsp; and&nbsp; $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$ become.
  
:* Die Distanz $D(P_X||P_Y)$ ist für $N = 1000$ größer als für $N = 400$.
 
Der Grund hierfür ist, dass das hier dokumentierte empirische Experiment mit $N = 400$ eher zu einer Gleichverteilung geführt hat als das Experiment mit $N = 1000$.
 
  
Würde man dagegen sehr (unendlich) viele Versuche mit $N = 400$ und $N = 1000$ starten und über diese mitteln, ergäbe sich tatsächlich der eigentlich erwartete monotone Verlauf.
+
However, one can also see from the table that there are exceptions:
 +
* The entropy&nbsp; $H(Y)$&nbsp; is smaller for&nbsp; $N = 1000$&nbsp; than for&nbsp; $N = 400$.
 +
* The distance&nbsp; $D(P_X\hspace{0.05cm}|| \hspace{0.05cm}P_Y)$&nbsp; is greater for&nbsp; $N = 1000$&nbsp; than for&nbsp; $N = 400$.
 +
* The reason for this is that the experiment documented here with&nbsp; $N = 400$&nbsp; was more likely to lead to a uniform distribution than the experiment with&nbsp; $N = 1000$.
 +
*If, on the other hand, one were to start a very (infinitely) large number of experiments with&nbsp; $N = 400$&nbsp; and&nbsp; $N = 1000$&nbsp; and average over all of them, the actually expected monotonic course would actually result.
  
 
{{ML-Fuß}}
 
{{ML-Fuß}}
Line 155: Line 186:
  
  
[[Category:Aufgaben zu Informationstheorie|^3.1 Vorbemerkungen zu 2D-Zufallsgrößen^]]
+
[[Category:Information Theory: Exercises|^3.1 General Information on 2D Random Variables^]]

Latest revision as of 10:14, 24 September 2021

Determined probability mass functions

The probability mass function is:

$$P_X(X) = \big[\hspace{0.03cm}0.25\hspace{0.03cm}, \hspace{0.15cm} 0.25\hspace{0.15cm},\hspace{0.15cm} 0.25 \hspace{0.03cm}, \hspace{0.15cm} 0.25\hspace{0.03cm}\big]\hspace{0.05cm}.$$

The random variable  $X$  is thus characterised by

  • the symbol set size  $M=4$,
  • equal probabilities  $P_X(1) = P_X(2) = P_X(3) = P_X(4) = 1/4$ .


The random variable  $Y$  is always an approximation for  $X$:

  • It was obtained by simulation from a uniform distribution, whereby only  $N$  random numbers were evaluated in each case.  This means:  
  • $P_Y(1)$, ... , $P_Y(4)$  are not probabilities in the conventional sense.  Rather, they describe  relative frequencies.


The result of the sixth test series  (with  $N=1000)$  is thus summarised by the following probability function:

$$P_Y(X) = \big [\hspace{0.05cm}0.225\hspace{0.15cm}, \hspace{0.05cm} 0.253\hspace{0.05cm},\hspace{0.15cm} 0.250 \hspace{0.05cm}, \hspace{0.15cm} 0.272\hspace{0.05cm}\big] \hspace{0.05cm}.$$

This notation already takes into account that the random variables  $X$  and  $Y$  are based on the same alphabet  $X = \{1,\ 2,\ 3,\ 4\}$.

With these preconditions, the  informational divergence  between the two probability functions  $P_X(.)$  and  $P_Y(.)$ :

$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = {\rm E}_X \hspace{-0.1cm}\left [ {\rm log}_2 \hspace{0.1cm} \frac{P_X(X)}{P_Y(X)}\right ] \hspace{0.2cm}=\hspace{0.2cm} \sum_{\mu = 1}^{M} P_X(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_X(\mu)}{P_Y(\mu)} \hspace{0.05cm}.$$

One calls  $D( P_X\hspace{0.05cm} || \hspace{0.05cm}P_Y)$  the (first)  Kullback-Leibler distance.

  • This is a measure of the similarity between the two probability mass functions  $P_X(.)$  and  $P_Y(.)$.
  • The expected value formation occurs here with regard to the  (actually equally distributed)  random variable  $X$.  This is indicated by the nomenclature  ${\rm E}_X\big[.\big]$.


A second form of Kullback-Leibler distance results from the formation of expected values with respect to the random variable  $Y$   ⇒   ${\rm E}_Y\big [.\big ]$:

$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) = {\rm E}_Y \hspace{-0.1cm} \left [ {\rm log}_2 \hspace{0.1cm} \frac{P_Y(X)}{P_X(X)}\right ] \hspace{0.2cm}=\hspace{0.2cm} \sum_{\mu = 1}^M P_Y(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_Y(\mu)}{P_X(\mu)} \hspace{0.05cm}.$$



Hints:


Questions

1

What is the entropy of the random variable  $X$ ?

$H(X)\ = \ $

$\ \rm bit$

2

What are the entropies of the random variables  $Y$  $($approximations for  $X)$?

$N=10^3\text{:} \hspace{0.5cm} H(Y) \ = \ $

$\ \rm bit$
$N=10^2\text{:} \hspace{0.5cm} H(Y) \ = \ $

$\ \rm bit$
$N=10^1\text{:} \hspace{0.5cm} H(Y) \ = \ $

$\ \rm bit$

3

Calculate the following Kullback-Leibler distances.

$N=10^3\text{:} \hspace{0.5cm} D( P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) \ = \ $

$\ \rm bit$
$N=10^2\text{:} \hspace{0.5cm} D( P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) \ = \ $

$\ \rm bit$
$N=10^1\text{:} \hspace{0.5cm} D( P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) \ = \ $

$\ \rm bit$

4

Does  $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$  give exactly the same result in each case?

Yes.
No.

5

Which statements are true for the Kullback-Leibler distances with  $N = 4$?

  $D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0$  is true.
  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0.5 \ \rm bit$  is true.
  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$  is infinitely large.
  $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0$  holds.
  $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0.5 \ \rm bit$  holds.
  $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$  is infinitely large.

6

Do both  $H(Y)$  and  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$  change monotonically with  $N$?

Yes,
No.


Solution

(1)  With equal probabilities, and with  $M = 4$:

$$H(X) = {\rm log}_2 \hspace{0.1cm} M \hspace{0.15cm} \underline {= 2\,{\rm (bit)}} \hspace{0.05cm}.$$


(2)  The probabilities for the empirically determined random variables  $Y$  generally  (not always!)  deviate from the uniform distribution the more the parameter  $N$  is smaller.  One obtains for

  • $N = 1000 \ \ \Rightarrow \ \ P_Y(Y) = \big [0.225, \ 0.253, \ 0.250, \ 0.272 \big ]$:
$$H(Y) = 0.225 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.225} + 0.253 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.253} + 0.250 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.250} + 0.272 \cdot {\rm log}_2 \hspace{0.1cm} \frac{1}{0.272} \hspace{0.15cm} \underline {= 1.9968\ {\rm (bit)}} \hspace{0.05cm},$$
  • $N = 100 \ \ \Rightarrow \ \ P_Y(Y) = \big[0.24, \ 0.16, \ 0.30, \ 0.30\big]$:
$$H(Y) = \hspace{0.05cm}\text{...} \hspace{0.15cm} \underline {= 1.9410\ {\rm (bit)}} \hspace{0.05cm},$$
  • $N = 10 \ \ \Rightarrow \ \ P_Y(Y) = \big[0.5, \ 0.1, \ 0.3, \ 0.1 \big]$:
$$H(Y) = \hspace{0.05cm}\text{...} \hspace{0.15cm} \underline {= 1.6855\ {\rm (bit)}} \hspace{0.05cm}.$$


(3)  The equation for the Kullback-Leibler distance we are looking for is:

$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \sum_{\mu = 1}^{4} P_X(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_X(\mu)}{P_Y(\mu)} = \frac{1/4}{{\rm lg} \hspace{0.1cm}(2)} \cdot \left [ {\rm lg} \hspace{0.1cm} \frac{0.25}{P_Y(1)} + \frac{0.25}{P_Y(2)} + \frac{0.25}{P_Y(3)} + \frac{0.25}{P_Y(4)} \right ] $$
$$\Rightarrow \hspace{0.3cm} D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot \left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{P_Y(1) \cdot P_Y(2)\cdot P_Y(3)\cdot P_Y(4)} \right ] \hspace{0.05cm}.$$

The logarithm to the base  $ 2$  ⇒   $\log_2(.)$  was replaced by the logarithm to the base  $ 10$   ⇒   $\lg(.)$ for easy use of the calculator.

The following numerical results are obtained:

  • for $N=1000$:
$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot \left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{0.225 \cdot 0.253\cdot 0.250\cdot 0.272} \right ] \hspace{0.15cm} \underline {= 0.00328 \,{\rm (bit)}} \hspace{0.05cm},$$
  • for $N=100$:
$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot \left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{0.24 \cdot 0.16\cdot 0.30\cdot 0.30} \right ] \hspace{0.15cm} \underline {= 0.0442 \,{\rm (bit)}} \hspace{0.05cm},$$
  • for $N=10$:
$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = \frac{1}{4 \cdot {\rm lg} \hspace{0.1cm}(2)} \cdot \left [ {\rm lg} \hspace{0.1cm} \frac{0.25^4}{0.5 \cdot 0.1\cdot 0.3\cdot 0.1} \right ] \hspace{0.15cm} \underline {= 0.345 \,{\rm (bit)}} \hspace{0.05cm}.$$


(4)  Correct is  No,  as will be shown by the example  $N = 100$ :

$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) = \sum_{\mu = 1}^M P_Y(\mu) \cdot {\rm log}_2 \hspace{0.1cm} \frac{P_Y(\mu)}{P_X(\mu)} = 0.24\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.24}{0.25} + 0.16\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.16}{0.25} +2 \cdot 0.30\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.30}{0.25} = 0.0407\ {\rm (bit)}\hspace{0.05cm}.$$
  • In subtask  (3)  we got  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0.0442$  instead.
  • This also means:   The designation „distance” is somewhat misleading.
  • According to this, one would actually expect  $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X) = D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$ .


Probability function, entropy and Kullback-Leibler distance

(5)  With  $P_Y(X) = \big [0, \ 0.25, \ 0.5, \ 0.25 \big ]$  one obtains:

$$D(P_X \hspace{0.05cm}|| \hspace{0.05cm} P_Y) = 0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0} + 2 \cdot 0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0.25}+0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0.50}\hspace{0.05cm}.$$
  • Because of the first term, the value of  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm}P_Y)$  is infinitely large.
  • For the second Kullback-Leibler distance holds:
$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0\cdot {\rm log}_2 \hspace{0.1cm} \frac{0}{0.25} + 2 \cdot 0.25\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.25}{0.25}+ 0.50\cdot {\rm log}_2 \hspace{0.1cm} \frac{0.5}{0.25} \hspace{0.05cm}.$$
  • After looking at the limits, one can see that the first term yields the result  $0$ .  The second term also yields zero, and one obtains as the final result:
$$D(P_Y \hspace{0.05cm}|| \hspace{0.05cm} P_X) = 0.50\cdot {\rm log}_2 \hspace{0.1cm} (2) \hspace{0.15cm} \underline {= 0.5\,{\rm (bit)}} \hspace{0.05cm}.$$

  Statements 3 and 5 are therefore correct:

  • From this extreme example it is clear that  $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$  is always different from  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$ .
  • Only for the special case  $P_Y \equiv P_X$  are both Kullback-Leibler distances equal, namely zero.
  • The adjacent table shows the complete result of this task.



(6)  Correct is again  No.   Although the tendency is clear:   The larger  $N$  is,

  • the more  $H(Y)$  approaches in principle the final value  $H(X) = 2 \ \rm bit$,
  • the smaller the distances  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm} P_Y)$  and  $D(P_Y\hspace{0.05cm}|| \hspace{0.05cm} P_X)$ become.


However, one can also see from the table that there are exceptions:

  • The entropy  $H(Y)$  is smaller for  $N = 1000$  than for  $N = 400$.
  • The distance  $D(P_X\hspace{0.05cm}|| \hspace{0.05cm}P_Y)$  is greater for  $N = 1000$  than for  $N = 400$.
  • The reason for this is that the experiment documented here with  $N = 400$  was more likely to lead to a uniform distribution than the experiment with  $N = 1000$.
  • If, on the other hand, one were to start a very (infinitely) large number of experiments with  $N = 400$  and  $N = 1000$  and average over all of them, the actually expected monotonic course would actually result.