Вы находитесь на странице: 1из 7

Population: a collection of all units of interest (people, products, )

e.g., all the 44 students in the classroom


Variables: a measurable property or attribute associated with each unit in the population.
e.g., interest in statistics, interest in programming
Parameters: numeric characteristics of the population defined for each variable of interest.
e.g., mean interest scores, standard deviation (STDEV) of interest scores
Sample: a subset of a population that is actually observed
e.g., 10 students randomly selected from the population (all the 44 students)

Random Variables (r.v.) X

e.g., Students' final score, quality of products, customers'


interarrival ti
How do we completely charaterize a r.v.?

Using its distribution.

What do we want to know about a r.v?

Population Parameters Sample Statistics

Mean Sample Mean

Standard Deviation Sample Stdev

Percentiles Sample
Percentile

The binomial distribution is the discrete probability distribution of


the number of successes in a sequence of n independent yes/no
experiments, each of which yields success with probability p.

Binominv(rand(), n, p)
Mean=Average(Array)
Standard Deviation=STDEV(Array)
A top 25%=75th Percentile=PERCENTILE(Array,K)=

PERCENTILE(B17:B76,0.75)
Total number of heads obtained= =SUM(Array)
Maximum of B4 and B5= =MAX(B4:B5)
Display "B4" if B4 > B5 otherwise display B5,

=IF(B4>B5,"b4","b5")
Outcome X =Rand() ; Returns an evenly distributed random
real number 0=< x <=1. A new random real number is
returned every time the worksheet is calculated.
X =Rand()*100, 0=< x <=100
BINOMINV= Inverse Cumulative Binomial Distribution
The Cumulative Binomial Distribution (calculated by the Excel
Binom.Dist function) is a statistical measure that is frequently
used to indicate the probability of obtaining a specific number
of successes from a specific number of independent trials.
The Binom.Inv function calculates the inverse of the
Cumulative Binomial Distribution. For a given number of
independent trials, the Binom.Inv function returns the smallest
value of x (the number of successes) for a specified Cumulative
Binomial Distribution probability.
The binomial distribution is the discrete probability distribution
of the number of successes in a sequence of n independent
yes/no experiments, each of which yields success with
probability p.
BINOMINV (probability of observing X successes in N trials, N independent
trials, P probability of success in each trial)

BINOMINV(RAND(),1,0.5)=If(Rand()<0.5,1,0) ; 1: Head, 0: Tail

If a donation is made, the amount of money that females


contribute follows a normal distribution with a mean of $20
and standard deviation of $3. The
NORMINV(probability,mean,standard_dev)=NORMINV(RAND(),2
0,3)
Index of coins Outcome X=DISCRINV(RAND(),N , P)=
=TRIANINV(Pro, Lower bond, Most likely, Upper bond) =TRIANINV(RAND(),0,1,1)
To create random variables with a Uniform distribution over the interval [ a, b ]
: a + RAND( )* (b a)
=EXPOINV(probability, mean)
POISINV(probability, mean)
Mod(10,4) = 2 gives the rise after division.

Possible Outcomes
Discrete, and a finite number of them.
# of workers absent out of 70 on a shift
number of defective parts in a lot of 100
number of available presses on Monday
Discrete, and an infinite number of them.
Annual demand for a product
Number of accidents during the year
Number of parts produced until the machine
Continuous, over a finite range.
The actual diameter of a machined part
Number of productive hours during a day
Interest rate change next year on an adjustable rate mortgage
Continuous, with fixed lower bound.
Total time to complete a work order
Time to failure of a machine or component
Continuous, with infinite upper and lower bounds.
hours a shipment is early or late
deviation of a machined diameter from specs
deviation of demand from forecast
COUNTIF(Array, Criteria) =COUNTIF(H31:H230,"0")
Percentile is indicating that the value below which a given
percentage which in our case is 95% that means all numbers
which are less or equal than 446 are included in this percentile.

Probability that the 95% percentile of the processing time of


these wafer lots
Number of interval time less than 446
=COUNTIF(B4:B103,"<=446")
Number of interval time bigger than 446
=COUNTIF(B4:B103,">446")
Sample Mean=AVERAGE(D10:D2009)
Sample Stdev=STDEV(D10:D2009)
CI for the mean? 95% Interval Exected
Lower Bound (LCL) =Mean-2*SD/SQRT(n)= $G$5-
2*$G$6/SQRT(2000)
Upper Bound (UCL)=Mean+2*SD/SQRT(n)
(UCL-LCL)/2 =(H17-H16)/2
PI for a particular outcome? 95% Interval Actual
Lower Bound (LPI)= Mean-2*SD*(1+1/SQRT(n))
=$G$5-2*$G$6*(1+1/SQRT(2000))
Upper Bound (UPI)= Mean+2*SD*(1+1/SQRT(n))

Random Rate of Return


Assets Value Now Rate of Return
Value Next Year
Bonds $5,000.00 =TRIANINV( RAND( ), 0.04, 0.06, 0.08 )
$5,334.18
Stocks $10,000.00 =NORMINV( RAND( ), 0.1, 0.1 )
$10,424.25
Cash $2,000.00 =RAND( ) *0.02+ 0.03
$2,095.83

Continuous Random Variables


Uniform=A+RAND()(B-A)
Exponential =EXPOINV(probability, mean)
Triangle =TRIANINV(RAND(),0,1,2)
Normal=NORMINV(Prob ,mean, STDEV) =NORMINV(RAND(),1,2)

Discrete Random Variables


Poisson=POISINV(probability, mean)
Bernoulli =BINOMINV(Prob,N,Prob) =BINOMINV(RAND(),1,0.9)
(Defective or nonfictive)
EX: There is a 40% chance that the drug will not be effective
=BINOMINV(RAND(),1,1-0.4)

Uniform Dist=RANDBETWEEN(bottom,top)=RANDBETWEEN(1,5)
What are the chances for me to have more than $18,500 in a
year?
=COUNTIF(D8:D2007,">=18500")/2000
Note:The variable cost per widget is approximately normally distributed with a mean of $55 and
standard deviation of $ 4, but the variable cost can never exceeds $65.
Variable Cost =MIN(NORMINV(RAND(),55,4),65)
Note:Sales to Little Guy are expected to remain the same as last year, but with an average
deviation of 15% either way. This years sales were $5.2 million. To represent this situation,
model next years sales using the distribution
NORMINV(RAND(),previous,0.15*previous)=NORMINV(RAND(),C6,0.15*C6)
Note:Zero-One Guy either places an order for $2 million per year, or orders nothing at all. This
year we did get an order for $2 million. Next year we think that there is a 60% chance he
will order; the year after we think the chance is 85%. Remember that this customer either places
an order for $2 million per year, or orders nothingat all. =2*BINOMINV(RAND(),1,0.6)
BINOMINVis used to generate a Bernoulli RV.) This indicates that there is a 60% chance
BINOMINVwill return a 1,and a 40% chance it will return a 0; the result will be multiplied by 2
million.

(a) Exponential:

(b) Poisson:

(c) Uniform:

(d) Normal:
Models the distribution of a process that can be thought of as the sum of a

(g)Triangle:
(f)
(f) Weibull:
Gamma:
Models
Models
An athe
extremelytime
process to failure
flexible
when (minimum
distribution
only ofto
the minimum,
used a number
model of possible
most likely
random causes);
andvariables
maximum withcan
valuesmodel
a lower
of the
bound.
(e) Lognormal:
Models the distribution of a process that can be thought of as the product of a number of

(g) Weibull:
Models the time to failure (minimum of a number of possible causes); can model

(f) Triangle:

Вам также может понравиться