Random Variables and Probability DistributionsRandom VariablesSlide 3Slide 4Slide 5Random Variable EventsProbability DistributionsProbability Mass FunctionSlide 9Probability Density FunctionSlide 11Slide 12Cumulative Distribution FunctionSlide 14Slide 15Slide 16Slide 17Joint Probability DistributionsMarginal Probability DistributionsDiscrete Marginal DistributionsContinuous Marginal DistributionsConditional DistributionsSlide 23Slide 24Slide 25Slide 26Slide 27IndependenceSlide 29Slide 301Random Variables andProbability DistributionsX is a random variable (r.v.)Probability Distribution•Discrete probability mass function (p.m.f.)•Continuous probability density function (p.d.f.)•Cumulative distribution function (c.d.f.)Joint Probability Distribution•Joint p.m.f or p.d.f.•Marginal distributions•Conditional distributions, Independence2Random VariablesX is a mapping from each simple event in the sample space S to a real number•Discrete: finite or countable set of values•Continuous: uncountable set of values3Random VariablesDiscrete random variable: only takes on certain values in a finite or countable set.•Flip two coins: X = # of heads HH X = 2 HT X = 1 TH X = 1 TT X = 0•Roll a die: X = # of dots on the side facing up•Color: X = # of students in a class whose favorite color is red•Post office: X = # of people in line4Random VariablesContinuous random variable: takes on any values within some range (i.e., infinitely many values in an uncountable set).•X = the lifetime (hours) of a light bulb•X = the weight of the next package that you take to the post office•X = the length of time to play 18 holes of golf•X = the annual income of Texas residents5Machine Breakdown Example S = {electrical, mechanical, misuse}Let X = the repair cost associated with a failureSample SpaceRandom Variable( Repair Cost )MisuseElectricalMechanicalReal Number Line50 200 350Random Variables6Random Variable EventsFlip two coins: X = # of heads•Exactly one head: [X = 1]•No more than one head: [X 1]Post office: X = # of people in line•At least 3 and fewer than 10 people: [3 X < 10]•More than 3 and at most 10 people: [3 < X 10]Light bulb: X = lifetime of the bulb•More than 50 hours: [X > 50]•Never turns on: [X = 0]7Probability DistributionsThe probabilities assigned to the values of XA discrete r.v. X has probability mass function (p.m.f.): fX(x) = P[X = x]A continuous r.v. X has probability density function (p.d.f.): fX(x)8Probability Mass FunctionfX(x) maps the possible values of the discrete r.v. X to probabilities on the interval [0,1].fX(x) = P[X = x] = probability that X = x, where x is from a finite or countable set.p.m.f. values are probabilities.Properties of the p.m.f.:1) 0 fX(x) 1 for all x2) x fX(x) = 19Probability Mass FunctionExample: Flip two coins•S = { HH, HT, TH, TT }•X = # of headsfX(0) = P[X = 0] = P(TT) = 1/4fX(1) = P[X = 1] = P(HT or TH) = 1/4 + 1/4 = 1/2fX(2) = P[X = 2] = P(HH) = 1/40 < fX(x) < 1 for x = 0, 1, 2; o/w fX(x) = 0x fX(x) = fX(0) + fX(1) + fX(2) = 110Probability Density FunctionfX(x) describes the distribution of values of X over a continuous range.p.d.f. values are NOT probabilities.Probabilities are found by calculating the area under the fX(x) curve.baXdxxfbXaP)(][11Probability Density FunctionProperties of the p.d.f.:1) fX(x) 0 for all x2)The probability of a single point is zeroIf X is a continuous r.v., then for any a and b,P[a X b] = P[a < X b] = P[a X < b] = P[a < X < b]1)( dxxfX0)(][ dxxfaXPaaX12Probability Density FunctionX = lifetime (hrs) of a certain kind of radio tubeP[X < 150] Note: 31100100150100100150100x0)( o/w ;100for 100)(2 xfxxxfXX110010001001001001002xdxxdxxdxxfX1501002150100)(13Cumulative Distribution FunctionFX(x) = P[X x] •Discrete r.v. X: for < x < •Continuous r.v. X: for < x < c.d.f. values are probabilitiesxXXdttfxF )()(xtXXtfxF )()(14Cumulative Distribution FunctionProperties of the c.d.f.:1) 0 FX(x) 1 for all x2) If x y, then FX(x) FX(y)15Cumulative Distribution FunctionDiscrete Example: Flip 2 coins•FX(0) = P[X 0] = fX(0) = 1/4•FX(1) = P[X 1] = fX(0) + fX(1) = 1/4 + 1/2 = 3/4•FX(2) = P[X 2] = fX(0) + fX(1) + fX(2) = 1 FX(x) = 0 for x < 0 FX(x) = 1/4 for 0 x < 1 FX(x) = 3/4 for 1 x < 2 FX(x) = 1 for x ≥ 216Machine Breakdown Example•X = the repair cost associated with a failure•P[X = 50] = 0.3, P[X = 200] = 0.2, P[X = 350] = 0.5CDF for XCumulative Distribution Function50 200 350Repair cost0.30.5117Cumulative Distribution FunctionContinuous Example: Lifetime of radio tube•For x > 100 FX(x) = 0 for x 100 FX(x) = for x > 100xxtdttdttfxFxxxXX1001100100100100100)()(1001002x100118Joint Probability DistributionsBivariate Case has two r.v.’s X and Y•fX,Y (x, y) is the joint p.m.f. or joint p.d.f.Joint p.m.f.: fX,Y (x, y) = P[X = x, Y = y]•0 fX,Y (x, y) 1 for all x, yx y fX,Y (x, y) = 1Joint p.d.f.:•fX,Y (x, y) 0 for all x, y• 1),(,dydxyxfYXdydxyxfAYXPAYX ),(]),[(,19Marginal Probability DistributionsFor discrete r.v.’s X and Y fX (x) = y fX,Y (x, y) fY (y) = x fX,Y (x, y) For continuous r.v.’s X and Y dxyxfyfdyyxfxfYXYYXX),()(),()(,,20Discrete Marginal DistributionsExample:21Continuous Marginal DistributionsExample: 0),( o/w ;0,0for 2),(,2,yxfyxeeyxfYXyxYX xxyxyxYXXeeeeeedyeedyyxfxf00202,222),()( yyxyxyYXYeeeeeedxeedxyxfyf2020202,22122),()(22Conditional DistributionsGiven fX,Y (x, y), fX (x) > 0, fY (y) > 0, we can calculate•fX|Y (x | y) = conditional distribution of X given Y = y•fY|X (y | x) = conditional distribution of Y given X = x)(),()|()(),()|(,|,|xfyxfxyfyfyxfyxfXYXXYYYXYX23Conditional DistributionsFor
View Full Document