Continuous Probability DistributionsContinuous Uniform DistributionSlide 3Slide 4Slide 5Normal DistributionSlide 7Standard Normal DistributionSlide 9Slide 10Slide 11Slide 12Slide 13Slide 14Slide 15Normal Distribution ProbabilitiesSlide 17Slide 18Slide 19Slide 20Normal Approx. to the BinomialSlide 22Slide 23Lognormal DistributionSlide 25Slide 26Exponential DistributionSlide 28Gamma DistributionSlide 30Slide 31Exponential and GammaSlide 33Slide 34Relationship to Poisson ProcessSlide 36Slide 37Chi-Squared (2) DistributionChi-Squared Distribution1Continuous Probability DistributionsContinuous Uniform DistributionNormal DistributionLognormal DistributionGamma and Exponential DistributionsChi-Squared Distribution2Continuous Uniform DistributionA continuous r.v. X with a “rectangular” p.d.f. over a specified range [A, B]•The p.d.f. and c.d.f depend on parameters A and BExpected ValueVariance 0)( o/w ;for 1),;( xfBxAABBAxfXX2][BAXE12)()(2ABXVBxAABAxxXPBAxFX for ][),;(3Continuous Uniform DistributionA BAB1xf(x)4Continuous Uniform DistributionExample: Harry is never early and can be up to 20 minutes late. Assume that his lateness follows a uniform distribution.•A = 0, B = 20410200151)20,0;15(1]15[ XFXPminutes 102200][ XE31001240012)020()(2XV5Continuous Uniform DistributionExample: Observe events occurring in a fixed time frame, e.g. buses arriving at a bus stop. •Suppose we know that one such event has occurred in the time interval (a, b).•X = the actual time of occurrence of the event.•A very simple model assumes that X is equally likely to lie in any small subinterval within (a, b). X follows a Uniform(a, b) distribution.6Normal DistributionA continuous r.v. X with a “bell curve” p.d.f. •The most important distribution in statistics•The p.d.f. depends on mean and variance 2Expected valueVarianceExcel function•p.d.f.: NORMDIST(x, , , 0)•c.d.f. : NORMDIST(x, , , 1)xexnxfor 21),;(221][XE2)(XV7Normal DistributionN(, 2)4 4x = 8Standard Normal DistributionLet X be distributed N(, 2)Then is distributed N(0, 1). •The p.d.f. for Z is•The c.d.f. for Z is tabulated in Table A.3zezzfor 21)(221XZ zzZPz for ][)(9Standard Normal DistributionHow to read Table A.3z .00 .01 .02 .030.0 0.5000 0.5040 0.5080 0.51200.1 0.5398 0.5438 0.5478 0.5517...1.2 0.8849 0.8869 0.8888 0.89071.3 0.9032 0.9049 0.9066 0.90820.9066)32.1(]32.1[ ZP0.09340.90661)32.1(1]32.1[ ZP5517.090490)13.0()31.1(]31.113.0[ .ZP10Standard Normal DistributionN(0, 1) (0) = 0.5 4 4x = 011Standard Normal Distribution P[Z 0] = P[Z 0] = 0.5 (0) = 0.50Z12Standard Normal DistributionP[1 Z 1] = P[Z 1] P[Z 1] = (1) (1) 1.0 1.01.01.013Standard Normal Distribution P[Z > z1] = 1 P[Z z1] = 1 (z1)0z114Standard Normal Distribution P[Z z1] = P[Z > z1] = 1 P[Z z1] (z1) = 1 (z1)z1Z015Standard Normal Distribution P[z1 Z z2] = P[Z z2] P[Z z1] = P[Z z2] (1 P[Z z1]) = (z2) (1 (z1) )z1z2016Normal Distribution ProbabilitiesLet X be distributed N(, 2)1) Standardize X2) Convert range for X[ ]a X bP a X b Pa bP Zm m ms s sm ms s- - -� �� � = � �� �� �- -� �= � �� �� �XZms-� =17Normal Distribution ProbabilitiesLet X be distributed N(, 2)3) Rewrite using standard normal c.d.f.4) Look up c.d.f. probabilities for z1 and z2 in Table A.3, then calculate.a b b aP Zm m m ms s s s- - - -� � � � � �� � =F - F� � � �� �� � � � � �bzaz21;z18Normal Distribution ProbabilitiesExample: The length of a part follows a normal distribution with mean 100mm and standard deviation 2mm.•X = length of the part ~ N( = 100, = 2)•What is the probability that the part length is above 103.3mm?[ ] [ ]( )103.3 100[ 103.3]21.65 1 1.651 1.65 1 0.9505 0.0495XP X PP Z P Zms- -� �> = >� �� �= > = - �= - F = - =19Normal Distribution ProbabilitiesExample (cont’d): length of a part.•X = length of the part ~ N( = 100, = 2)•What proportion of the part output will have lengths between 98.5mm and 102mm?[ ]( ) ( ) ( ) ( )98.5 100 102 100[98.5 102]2 20.75 1.01.0 0.75 1.0 1 0.750.8413 (1 0.7734) 0.6147XP X PP Zms- - -� �� � = � �� �� �= - � �� �=F - F - =F - - F� �= - - =20Normal Distribution ProbabilitiesExample: Suppose Harry’s lateness is normally distributed with a mean of 2 minutes and a variance of 16.What is the probability that Harry will be between 1 and 5 minutes late?•X = Harry’s lateness ~ N( = 2, 2 = 16) 3721.0)5987.01(7734.025.075.075.025.0425421]51[ZPXPXP21Normal Approx. to the BinomialX follows a Bin(n, p) distributionIf n is large, np > 5 and n(1p) > 5, then is approximately N(0, 1))1()5.0()1()5.0()1()5.0()1()5.0(),;(pnpnpxpnpnpxpnpnpxZpnpnpxPpnxb)1( pnpnpXZ22Normal Approx. to the BinomialExample: n = 10, p = 0.523Normal Approx. to the BinomialExample: Flip a coin 40 times.•X = # heads ~ Bin(n = 40, p = 0.5)•np = (40)(0.5) = 20; np(1p ) = (20)(0.5) = 10 7074.0)9222.01(7852.042.1179.079.042.110205.2210205.15708275.0)5.0,40;15()5.0,40;22(]2216[ZPZPBBXP24Lognormal DistributionA continuous r.v. X such that r.v. Y = ln(X) is normally distributed•Y has mean and variance 2•The p.d.f. depends on parameters and Expected valueVariance0)( o/w ;0for
View Full Document