1 special continuous probability distributions -normal distributions -lognormal distributions dr....
TRANSCRIPT
1
Special Continuous Probability Distributions-Normal Distributions
-Lognormal Distributions
Dr. Jerrell T. Stracener, SAE Fellow
Leadership in Engineering
EMIS 7370/5370 STAT 5340 : PROBABILITY AND STATISTICS FOR SCIENTISTS AND ENGINEERS
Systems Engineering ProgramDepartment of Engineering Management, Information and Systems
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
2
A random variable X is said to have a normal (orGaussian) distribution with parameters and ,where - < < and > 0, with probability density function
for - < x <
222
1
2
1)(
x
exf
f(x)
x
Normal Distribution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
3
the effects of and
Properties of the Normal Model
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
4
• Mean or expected value ofMean = E(X) =
• Median value of
X0.5 =
• Standard deviation
)(XVar
X
X
Normal Distribution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
5
Standard Normal Distribution
• If ~ N(, )
and if
then Z ~ N(0, 1).
• A normal distribution with = 0 and = 1, is calledthe standard normal distribution.
X
Z
X
Normal Distribution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
6
x 0 z
σ
μx'Z
f(x) f(z)
P (X<x’) = P (Z<z’)
X’ Z’
Normal Distribution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
7
• Standard Normal Distribution Table of Probabilities
http://www.engr.smu.edu/~jerrells/courses/help/normaltable.html
Enter table with
and find thevalue of
• Excelz
0
z
f(z)
x
Z
Normal Distribution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
8
The following example illustrates every possible case of application of the normal distribution.
Let ~ N(100, 10)
Find:
(a) P(X < 105.3)
(b) P(X 91.7)
(c) P(87.1 < 115.7)
(d) the value of x for which P( x) = 0.05X
X
X
Normal Distribution - Example
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
9
a. P( < 105.3) =
= P( < 0.53)= F(0.53)= 0.7019
10
1003.105P
X
100 x 0 z
f(x) f(z)
105.3 0.53
X
Z
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
10
b. P( 91.7) =
= P( -0.83) = 1 - P( < -0.83) = 1- F(-0.83)
= 1 - 0.2033 = 0.7967
10
1007.91
X
P
100 x 0 z
f(x) f(z)
91.7 -0.83
ZZ
X
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
11
c. P(87.1 < 115.7) = F(115.7) - F(87.1)
= P(-1.29 < Z < 1.57)= F(1.57) - F(-1.29)= 0.9418 - 0.0985 = 0.8433
7.115
10
1001.87
x
P
100
x
f(x)
87.1 115.7 0
x
f(x)
-1.29 1.57
X
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
12
100x
0z
f(x) f(z)
10
10064.1
x
0.05 0.05
1.64116.4
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
13
(d) P( x) = 0.05P( z) = 0.05 implies that z = 1.64P( x) =
therefore
x - 100 = 16.4x = 116.4
10
1001
10
100P
10
100P
xxZ
xX
64.110
100
x
ZX
X
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
14
The time it takes a driver to react to the brake lightson a decelerating vehicle is critical in helping toavoid rear-end collisions. The article ‘Fast-Rise BrakeLamp as a Collision-Prevention Device’ suggests that reaction time for an in-traffic response to abrake signal from standard brake lights can be modeled with a normal distribution having meanvalue 1.25 sec and standard deviation 0.46 sec.What is the probability that reaction time is between1.00 and 1.75 seconds? If we view 2 seconds as acritically long reaction time, what is the probabilitythat actual reaction time will exceed this value?
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
15
75.100.1 XP
46.0
25.175.1
46.0
25.100.1XP
09.154.0 XP
54.009.1 FF
5675.02946.08621.0
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
16
2XP
0516.0
9484.01
63.11
63.1
46.0
25.12
F
ZP
ZP
Normal Distribution – Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
17
Statistical Quality Control is an application of probabilitistic and statistical techniques to quality control
An Application of Probability & Statistics Statistical Quality Control
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
18
Statistical Quality Control - Elements
Analysisof processcapability
Statisticalprocesscontrol
Processimprovement
Acceptancesampling
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
19
Statistical Tolerancing - Convention
Normal ProbabilityDistribution
LTL Nominal UTL
0.001350.00135 0.9973
+3-3
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
20
Statistical Tolerancing - Concept
LTL UTLNominal x
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
21
For a normal distribution, the natural tolerance limits include 99.73% of the variable, or put another way, only 0.27% of the process output will fall outside the natural tolerance limits. Two points should be remembered:
1. 0.27% outside the natural tolerances sounds small, but this corresponds to 2700 nonconforming parts per million.
2. If the distribution of process output is non normal, then the percentage of output falling outside 3 may differ considerably from 0.27%.
Caution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
22
The diameter of a metal shaft used in a disk-drive unitis normally distributed with mean 0.2508 inches andstandard deviation 0.0005 inches. The specificationson the shaft have been established as 0.2500 0.0015 inches. We wish to determine what fraction ofthe shafts produced conform to specifications.
Normal Distribution - Example
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
23
spec meetingP
91924.0
0000.091924.0
60.41.40
0005.0
0.2508-0.2485
0.0005
0.2508-0.2515
x0.2485P0.2515xP
0.2515x0.2485P
0.2508 0.2515 USL
0.2485 LSL
0.2500
f(x)
xnominal
Normal Distribution - Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
24
Thus, we would expect the process yield to be approximately 91.92%; that is, about 91.92% of the shafts produced conform to specifications. Note that almost all of the nonconforming shafts are too large, because the process mean is located very near to the upper specification limit. Suppose we can recenter the manufacturing process, perhaps by adjusting the machine, so that the process mean is exactly equal to the nominal value of 0.2500. Then we have
Normal Distribution - Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
25
0.2515x0.2485P
9973.0
00135.099865.0
00.33.00
0005.0
0.2500-0.2485
0.0005
0.2500-0.2515
x0.2485P0.2515xP
0.2500 0.2515 USL
0.2485 LSL
f(x)
xnominal
Normal Distribution - Example Solution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
26
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
27
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
28
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
29
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
30
Using a normal probability distribution as a model for a quality characteristic with the specification limits at three standard deviations on either side of the mean. Now it turns out that in this situation the probability of producing a product within these specifications is 0.9973, which corresponds to 2700 parts per million (ppm) defective. This is referred to as three-sigma quality performance, and it actually sounds pretty good. However, suppose we have a product that consists of an assembly of 100 components or parts and all 100 parts must be non-defective for the product to function satisfactorily.
Normal Distribution - Example
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
31
The probability that any specific unit of product is non-defective is
0.9973 x 0.9973 x . . . x 0.9973 = (0.9973)100 = 0.7631
That is, about 23.7% of the products produced underthree sigma quality will be defective. This is not anacceptable situation, because many high technology products are made up of thousands of components.An automobile has about 200,000 components andan airplane has several million!
Normal Distribution - Example
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
32
Lognormal Distribution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
33
Definition - A random variable is said to have the Lognormal Distribution with parameters and , where > 0 and > 0, if the probability density function of X is:
, for x > 0
, for x 0
22
xln2
1
e2x
1 )x(f
0
x
f(x)
0
X
Lognormal Distribution
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
34
• Rule: If ~ LN(,),
then = ln ( ) ~ N(,)
• Probability Distribution Function
where F(z) is the cumulative probability distribution function of N(0,1)
xFxF
ln )(
Y
X
X
Lognormal Distribution - Properties
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
35
Mean or Expected Value
22
1
)(
eXE
2
1
12σe
2σ2μeSD(X)
• Standard Deviation
• Median
ex 5.0
Lognormal Distribution - Properties
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
36
A theoretical justification based on a certain materialfailure mechanism underlies the assumption that ductile strength X of a material has a lognormal distribution. Suppose the parameters are = 5 and = 0.1
(a) Compute E( ) and Var( )(b) Compute P( > 120)(c) Compute P(110 130)(d) What is the value of median ductile strength?(e) If ten different samples of an alloy steel of this type were subjected to a strength test, how many would you expect to have strength at least 120?(f) If the smallest 5% of strength values were unacceptable, what would the minimum acceptable strength be?
XX
XX
Lognormal Distribution - Example
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
37
Lognormal Distribution –Example Solution
a)
223)1()(
16.149)(22
2
2
005.5005.52
eeXVar
eeeXEu
u
b) )120(1)120( XPXP
9834.0
0166.01
)13.2(1
)1.0
0.5120ln(1
F
ZP
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
38
Lognormal Distribution –Example Solution
c) )1.0
0.5130ln
1.0
0.5110ln()130110(
ZPXP
092.0
0014.00934.0
)99.2()32.1(
)32.199.2(
FF
ZP
d) 41.14855.0 eemedianX u
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
39
Lognormal Distribution –Example Solution
e) )120( XPP
983.0
0170.01
)12.2(1
)1.0
0.5120ln(1
)120(1
F
ZP
XP
Let Y=number of items tested that have strength of at least 120Y=0,1,2,…,10
Stracener_EMIS 7370/STAT 5340_Sum 08_06.05.08
40
Lognormal Distribution –Example Solution
83.9983.0*10)(
)983.0,10(~
npYE
BY
f) The value of x, say xms, for which is determined as follows:
05.0)( msxXP
964.125
64.11.0
0.5ln
05.0)64.1(
05.0)1.0
0.5ln(
ms
ms
ms
x
x
ZP
xZP