25.08.2021 Views

082-Engineering-Mathematics-Anthony-Croft-Robert-Davison-Martin-Hargreaves-James-Flint-Edisi-5-2017

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

964 Chapter 29 Statistics and probability distributions

N(x)

N(x)

N(x) (a) (b)

(a)

m

x

(b)

m

x

m 1 m 2 x

Figure29.10

Two typical normal curves.

Figure29.11

Two normalcurveswith the same

standarddeviationbutdifferentmeans.

Letxbeacontinuous random variable with a normal distribution,N(x). Then

N(x) = 1

σ √ 2π e−(x−µ)2 /2σ 2

−∞<x<∞

where µ =expected(mean)valueofx, σ =standarddeviationofx.Figure29.10shows

two typical normal curves. All normal distributions are bell shaped and symmetrical

about µ.

In Figure 29.10(a) the values ofxare grouped very closely to the mean. Such a distribution

has a low standard deviation. Conversely, in Figure 29.10(b) the values of the

variable are spread widely about the mean and so the distribution has a high standard

deviation.

Figure29.11showstwonormaldistributions.Theyhavethesamestandarddeviation

butdifferentmeans.ThemeanofthedistributioninFigure29.11(a)is µ 1

whilethemean

of that in Figure 29.11(b) is µ 2

. Note that the domain ofN(x) is (−∞,∞); that is, the

domain is all real numbers. As for all distribution curves the total area under the curve

is1.

29.14.1 Thestandardnormal

A normal distribution is determined uniquely by specifying the mean and standard deviation.

The probability thatxliesinthe interval [a,b] is

P(axb)=

∫ b

a

N(x)dx

Themathematicalformofthenormaldistributionmakesanalyticintegrationimpossible,

so all probabilities must be computed numerically. As these numerical values would

change every time the value of µ or σ was altered some standardization is required. To

this end we introduce the standard normal. The standard normal has a mean of 0 and

a standard deviation of 1.

Considertheprobabilitythattherandomvariable,x,hasavaluelessthanz.Forconvenience

wecall thisA(z).

A(z) =P(x <z) =

∫ z

−∞

N(x)dx

Figure29.12illustratesA(z).ValuesofA(z)havebeencomputednumericallyandtabulated.TheyaregiveninTable29.7.Usingthetableandthesymmetricalpropertyofthe

distribution,probabilities can becalculated.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!