4 Kontinue Waarskynlikheids

Verdelings
Continuous Probability
Distributions
Did you hear about the statistician who had his head in an oven
and his feet in a bucket of ice? When asked how he felt, he
replied, "On the average I feel just fine."

It is proven that the celebration of birthdays is healthy. Statistics

show that those people who celebrate the most birthdays
become the oldest.

Kontinueverdelings
Continuous Distributions

Die veranderlikes waarna ons in hierdie hoofstuk gaan kyk

is nou kontinue veranderlikes. Dit beteken dat inplaas
daarvan om gebeurtenisse te tel, ons nou hoeveelhede
het wat gemeet gaan word.
The variables that we shall be looking at in this chapter will
be continuous variables. This means that instead of
counting the number of events, we shall have quantities
that we are measuring.

Vir diskrete verdelings het ons waarskynlikhede bepaal

deur (X = x) te bereken en dan, indien nodig, om
verskeie hiervan bymaak te tel.
For discrete distributions we calculated probabilities by
calculating (X = x) and then if it was necessary, we
added =‘s together.

0 1 2 3 4 5

1
By kontinue verdelings word waarskynlikhede bereken deur
oppervlaktes te bepaal.
For continuous distributions probabilities are calculated by
determining areas.
0.6

P(X > 1) = 0.5

0.4

0.3

0.2

0.1

0
-3

3
-2.7
-2.4
-2.1
-1.8
-1.5
-1.2
-0.9
-0.6
-0.3

0.3
0.6
0.9
1.2
1.5
1.8
2.1
2.4
2.7

P(X = x) = 0 vir enige kontinue verdeling

for any continuous distribution
P(X ≥ x) = P(X > x) en/and P(X ≤ x) = P(X < x)

Waarskynlikheidsdigtheidsfunksie
Probability density function

Kumulatiewe verdelingsfunksie
Cumulative distribution function

EXAMPLE 4-2 (p. 110)

EXAMPLE 4-4 (p. 112)

Verwagte Waarde
Expected Value

µ≡ Gemiddelde van ‘n waarskynlikheidsverdeling

Mean of a probability distribution

Gemiddeld, wat verwag ons gaan gebeur.

On average, what do we expect to happen.

2
Variansie en Standaard Afwyking
Variance and Standard Deviation

σ2 ≡ Variansie van waarskynlikheidsverdeling

Variance of probability distribution

σ≡ Standaard Afwyking van waarskynlikheidsverdeling

Standard Deviation of probability distribution

EXAMPLE 4-8 (p. 115)

Kontinue Uniforme Verdeling

Continuous Uniform Distribution

EXAMPLE 4-9 (p. 116)

Normaal Verdeling
Normal Distribution
Een van die belangrikste verdelings in statistiek.
One of the most important distributions in statistics.

Parameters:
µ: Populasie Gemiddelde, E(X)
Population Mean, E(X)
σ: Populasie Standaard Afwyking
Population Standard Deviation

X ~ Nor(µ, σ2)

3
Klokvormig / Bell Shaped

Wiskundig word oppervlaktes bereken deur funksies te

integreer. Ons moet dus die normaal funksie integreer
om verskillende waarskynlikhede te kry.
Mathematically areas are calculated by integrating
functions. We should thus integrate the normal function
to calculate desired probabilities.

Dit is baie tydrowend en geensins maklik nie.

This is time-consuming and not too easy.

Gelukkig is daar tabelle beskikbaar vanwaar ons

waarskynlikhede kan aflees.
Luckily normal tables are available from where we can read
off probabilities.

Transformasie
Transformation

Elke kombinasie van µ en σ gee vir ons ‘n ander normaal

verdeling.
Each combination of µ en σ gives a different normal
distribution.

Gemiddelde gewig: Mans = 75, Vrouens = 65

Average weight: Males = 75, Females = 65

Gemiddelde Inkomste: R14 000 of R 14 900

Average Income:

4
Gemiddelde gewig: Mans = 75, Vrouens = 65
Average weight: Males = 75, Females = 65
Gewig Mans en Dames / Weight Males and Females

Gewig (kg) / Weight (kg)

Ons sal dus ‘n verskillende tabel vir elke moontlike

kombinasie van µ en σ nodig hê.
We would need a different table for every possible
combination of µ and σ.

Gelukkig bestaan daar ‘n eenvoudige transformasie wat

gebruik kan word om enige normaal verdeling te
transformeer na ‘n standaard normaal vedeling.
Luckily an easy transformation exists that can be used to
transform any normal distribution into a standard normal
distribution.

Z ~ Nor(0, 1)
Tabel / Table III

Gee slegs P(Z < z), dus alle vrae moet geskryf word in
terme van <.
Only gives P(Z < z), thus all questions must be written in
terms of <.

5
Berekening van Normaal waarskynlikhede
Calculating Normal Probabilities

1) X  Z

2) P(Z < a | Z > b) = P(b < Z < a) / P(Z > b)

3) P(a < Z < b) = P(Z < b) – P(Z < a)

4) P(Z > a) = 1 – P(Z < a)

5) Tabel
Table

EXAMPLE 4-13 (p. 122)

Vanaf ‘n Waarskynlikheid na ‘n X-Waarde

From a Probability to an X-Value

EXAMPLE 4-14 (p. 123)

6
1) X  Z

2) Z  p  Kry waarskynlikheid
Get probability

Nou / Now:
Het waarskynlikheid, soek X-waarde
We have a probability, we’re looking for an X-value

1) Waarskynlikheid  PZ
Probability

2) Z  X

Normaalbenaderings
Normal Approximations

Binomiaalverdeling
Binomial Distribution

Indien kan die

normaalverdeling gebruik word om binomiaal
waarslynlikhede te bereken.
If then the normal distribution
van be used to calculate binomial probabilities.

Kontinuïteitskorreksie / Continuity correction

EXAMPLE 4-17 (p. 128)

EXAMPLE 4-18 (p. 129)
EXAMPLE 4-19 (p. 129)

7
Poissonverdeling
Poisson Distribution

Indien kan die normaalverdeling gebruik word om

Poisson waarslynlikhede te bereken.
If then the normal distribution van be used to
calculate Poissonbinomial probabilities.

Kontinuïteitskorreksie / Continuity correction

EXAMPLE 4-20 (p. 130)

Eksponentiaalverdeling
Exponential Distribution

Nuttige verdeling indien ons belangstel in die tyd totdat ‘n

gebeurtenis plaasvind.
Useful distribution if we are interested in the time until an
event occurs.

Parameter:
λ : Gemiddelde aantal gebeurtenisse per eenheid.
Average number of events per unit.

EXAMPLE 4-21 (p. 133)

EXAMPLE 4-22 (p. 135)

8
 Erlang en Gamma Verdelings
Erlang and Gamma Distributions

Nuttige verdeling indien ons belangstel in die tyd totdat die

r-de gebeurtenis plaasvind.
Useful distribution if we are interested in the time until the
r-th event occurs.

Erlang  r heelgetal / integer,

Gamma  r > 0

Onvolledige gammafunksie
Incomplete gamma function

Gebruik Poisson of Erlang/gamma direk.

Use Poisson or Erlang/gamma directly.

Die onvolledige gammafunksie: The incomplete gamma function

Area na links – Area to the left

x\ α=r 1 2 3 4 5 6 7 8 9 10 11 12
1 0.6321 0.2642 0.0803 0.0190 0.0037 0.0006 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000
2 0.8647 0.5940 0.3233 0.1429 0.0527 0.0166 0.0045 0.0011 0.0002 0.0000 0.0000 0.0000
3 0.9502 0.8009 0.5768 0.3528 0.1847 0.0839 0.0335 0.0119 0.0038 0.0011 0.0003 0.0001
4 0.9817 0.9084 0.7619 0.5665 0.3712 0.2149 0.1107 0.0511 0.0214 0.0081 0.0028 0.0009
5 0.9933 0.9596 0.8753 0.7350 0.5595 0.3840 0.2378 0.1334 0.0681 0.0318 0.0137 0.0055
6 0.9975 0.9826 0.9380 0.8488 0.7149 0.5543 0.3937 0.2560 0.1528 0.0839 0.0426 0.0201
7 0.9991 0.9927 0.9704 0.9182 0.8270 0.6993 0.5503 0.4013 0.2709 0.1695 0.0985 0.0533
8 0.9997 0.9970 0.9862 0.9576 0.9004 0.8088 0.6866 0.5470 0.4075 0.2834 0.1841 0.1119
9 0.9999 0.9988 0.9938 0.9788 0.9450 0.8843 0.7932 0.6761 0.5443 0.4126 0.2940 0.1970
10 1.0000 0.9995 0.9972 0.9897 0.9707 0.9329 0.8699 0.7798 0.6672 0.5421 0.4170 0.3032
11 1.0000 0.9998 0.9988 0.9951 0.9849 0.9625 0.9214 0.8568 0.7680 0.6595 0.5401 0.4207
12 1.0000 0.9999 0.9995 0.9977 0.9924 0.9797 0.9542 0.9105 0.8450 0.7576 0.6528 0.5384
13 1.0000 1.0000 0.9998 0.9989 0.9963 0.9893 0.9741 0.9460 0.9002 0.8342 0.7483 0.6468
14 1.0000 1.0000 0.9999 0.9995 0.9982 0.9945 0.9858 0.9684 0.9379 0.8906 0.8243 0.7400
15 1.0000 1.0000 1.0000 0.9998 0.9991 0.9972 0.9924 0.9820 0.9626 0.9301 0.8815 0.8152
16 1.0000 1.0000 1.0000 0.9999 0.9996 0.9986 0.9960 0.9900 0.9780 0.9567 0.9226 0.8730
17 1.0000 1.0000 1.0000 1.0000 0.9998 0.9993 0.9979 0.9946 0.9874 0.9739 0.9509 0.9153
18 1.0000 1.0000 1.0000 1.0000 0.9999 0.9997 0.9990 0.9971 0.9929 0.9846 0.9696 0.9451
19 1.0000 1.0000 1.0000 1.0000 1.0000 0.9998 0.9995 0.9985 0.9961 0.9911 0.9817 0.9653
20 1.0000 1.0000 1.0000 1.0000 1.0000 0.9999 0.9997 0.9992 0.9979 0.9950 0.9892 0.9786

9
λ = 0.001 per uur / per hour

X ≡ Tyd tot die 4de faling.

Time till the 4th failure.

P(X > 2000) = ??  Erlang

P(N ≤ 3 | 2000 uur)  Poisson

EXAMPLE 4-24 (p. 140)

Weibullverdeling
Weibull Distributions

Tyd tot faling.

Time to failure.

EXAPMLE 4-25 (p. 143)

10
Lognormaalverdeling
Lognormal Distributions

W ~ Nor(μW,σ2W)

X = exp(W)

X ~ lognormal(μX,σ2X)

EXAMPLE 4-26 (p. 145)

Betaverdeling
Beta Distributions

Begrens tussen [0 ; 1]
Bounded between [0 ; 1]

EXAMPLE 4-27 (p. 147)

EXAMPLE 4-28 (p. 147)

11