Apartment complex data Apartment] Floor | No. of | Size of | Distance of apart- number | number | bedrooms | apartment | ment from lift (sat) | (meters) ivy fa T 900.23 | 503.5 2 i 2 1175.34 | 325.6 3 i 3 1785.85 | 4508 4 2 1 900.48 | 500.1 5 2 2 1175.23 | 3245 6 2 3 1785.35 | 456.7 7 3 1 90053 | 502.5 [3_ [3 [2 1176.34 | 325.6 9 3 3 1787.85 | 450.8 10 4 1 900.78 | 500.1 i 4 2 1176.03 _| 325.4 R 4 3 1784.85. | 455.7 Questions 1. Let the random variable be number of bedrooms, what are the possible values that might be observed? Answer: 1,2,3 2. Let the random variable be floor number of the apartment. What are the possible values that might be observed? Answer: 1, 2,3,4 3. Let the random variable be size of the apartment. What are the possible values that might be observed? Answer: [900,180] sq. ft 4. Let the random variable be distance of the apartment from the lift. What are the possible values that might be observed? Answer: [324,505] meters Probability Mass Functions (p.m.f.) Suppose there are n random variables. There are two condi function: 1) P(X=xi) > Oe. it cannot be negative 2) The summation of all n random variables should add up to 1 ns to satisfy a probability massProperties of p.m.f > The probability mass function p(x) is positive for at most a countable number of values of x. That is, if X must assume ‘one of the values x1, x2, then 1. pla) 20,7 =1,2... 2. p(x) =0 for all other values of x > Represent it in tabular form x x E3 x P(X = x) | pCa) | pba) | pbs) > Since X must take one of the values x, we have seta) ed iss am Example 1 (for a finite set of discrete random variables): > Suppose X is a random variable that takes three values, 0, 1, and 2 with probabilities > (0) = P(X =0)= 4 > pl) =P(X=1)= 4 1d20 ree)=Px=2)=f > Tabular form eee x 1/2 ad iala} oS” lala : ay aan POP Woy 7 ow bad ted eta 4 0. ee a Wooe 78 tas Example 2 (for an infinite set of disorete random variables):> Suppose X is a random variable that takes values, 0, 1, 2, ..with probabilities x > pli) = cA, for some positive A > What is the value of c? > Se) =1 yt g or Graphing the p.m. Example: positive skewed distribution x Tess PR=a) Loa [03 [02 | Or Probability Note - You can have uniform, skewed and symmetric graphs. You can also have graphs that do not indicate any clear pattern.Rolling a dice twice: X-sum of outcomes x 273 Pea) TE Ti troll ke ke H = 3 a 3 02 Rolling a dice twice: X-smaller of outcomes 73 T4] 5] 6 T Eletetetete) 04 03 02 7 | [| ° Eo 1 2 3 ‘ . ° x Cumulative Distribution Function (c.d.f.) a é Here, the random variable X takes the value of P(xi) cumulatively. Thi similar to a step function,> The cumulative distribution function (cdf), F, can be expressed by F(a) = P(X < a) > If X isa discrete random variable whose possible values are 1,90, where Xt < x2 <%5..., then the distribution function F of X is a step function. Step function > Let X be a discrete random variable with the following probability mass function. food | > The cumulative distribution function of X is given by 0 3 lsa<2 3 a Fla)= we A o A * Until a=1, F(a) is treated as 0. Then from 1 to 2 for a, F(a) becomes 0.5.Note: Here, X or F(a) is a discrete random variable (not continuous). ‘When you need to solve for P(X=4) as shown below, you need to subtract upper value from the lower value Week 10Expectation of a Random variable Expected value of a random variable is the weighted average of the sum of relative frequencies taken in the long run of a repeated event. Expectation of a random variable Definition Let X be a discrete random variable taking values x1,%2 The expected value of X denoted by E(X) and referred to as Expectation of X is given by E(x) > xiP(X =x) = Rolling a dice once > > > > v Random experiment: Roll a dice once. Sample space: S = {1,2,3,4,5,6} Random variable X is the outcome of the roll. The probability distribution is given by x T[2];3]4][s5]6 PRER E(X) = 1§ + 25 +35 +43 +53 +63 = 3.5. > Does this mean that if we roll a dice once, should we expect the outcome to be 3.5? NOI!-the expected value tells us is what we would expect the average of a large number of rolls to be in the long run. Bernoulli random variableBernoulli random variable > A random variable that takes on either the value 1 or 0 is called a Bernoulli random variable. > Let X be a Bernoulli random variable that takes on the value 1 with probability p. > The probability distribution of the random variable is [x o [1 P(X = x) a) > Expected value of a Bernoulli random varaible: E(X) =0x(1—p)+1xp=p Discrete uniform random variable Discrete uniform random variable > Let X be a random variable that is equally likely to takes any of the values 1,2... > Probability mass function x P(X =x) : A gery a E(x)= ha Gb -- e| uae -- n omeProperties of Expectation of a Random variable Expectation of a function of a random variable Proposition Let X be a discrete random variable which takes values x; along with its probability mass function, P(X = x;). Let g be any real values function, The expected value of g(X) is E(@(X)) = Yo as) P(X = x) Corollary Ifa and b are constants, E(aX + b) = aE(X) +b Example > Let X be a discrete random variable with the following distribution TOT2] cu-07t03 J) 02/05] 03) > Let ¥ = g(X) = X*, What is E(Y)? > E(Y) =(-1) x02+0%05+17x03=05 > Distribution of Y ¥ Tey] PY [os;os ew # FOVariance of a random variable Variance of a random variable > Let's denote expected value of a random variable X by the greek alphabet Definition Let X be a random variable with expected value j1, then the variance of X, denoted by Var(X) or V(X), is defined by Var(X) = E(X — p)? > In other words, the Variance of a random variable X measures the square of the difference of the random variable from its ‘mean, j1, on the average. Computational formula for Var(X) y= & bet Gen We poo FX Eloxrk) = ableItb clan E (Rr 2p) = EW =F 7p r i. seep ep = OOFBernoulli random variable > A random variable that takes on either the value 1 or 0 is called a Bernoulli random variable. » Let X be a Bernoulli random variable that takes on the value 1 with probability p. > The probability distribution of the random variable is 0 [a {2 1-p|p] > Expected value of a Bernoulli random varaible: E(X)=0x(1—p) +1xp=p > Variance of a Bernoulli random variable: Var(X) = p — p? = p(1— p), Discrete uniform random variable > Let X be a random variable that is equally likely to takes any of the values 1,2,.. > Probability mass function x T][2 x 1/4 P(X Wt > E(X) = NY phi Discrete uniform random variable > Let X be a random variable that is equally likely to takes any of the values 1,2,..., > Probability mass function Hala 5 E(x) = (x2) = (enn Var(X), -G@) vvyProperties of Variance of expectation Variance of a function of a random variable Proposition Let X be a random variable, let c be a constant, then > Var(eX) = eVar(X) > Var(X +.) = Var(X) Corollary Ifa and b are constants, V(aX + b) E(aX + But if events are independent, then, Variance of sum of independent random variables Result Let X and Y be independent random variables. Then Var(X + ¥) = Var(X) + Var Y) ————————— eetVariance of sum of many independent random variables > The result that the variance of the sum of independent random variables is equal to the sum of the variances holds for not only two but any number of random variables. > Let X1,X2,...,X be k discrete random variables. Then, Standard Deviation (SD) of expectation SD of an expectation is the square root of the variance of the expectation. Note- Here, expectation does not reflect risk but the standard deviation reflects risk and this is why sd is important. Refer to this video for a detailed explanation on its application. Properties of standard deviation Proposition Let X be a random variable, let c be a constant, then > SD(cX) = SD(X) > SD(X +¢) = SD(X) 1. If Var(X) = 4, what is SD(3X)? Answer: 6. 2. If Var(2X +3) = 16, what is SD(X)? Answer: 2. Important ProblemsWeek 11 Binomial distribution - Bernoulli distribution ‘ABemoulli trial has a sample set of two outcomes, namely, success and failure. Examples of Bernoulli trials > Experiment: Tossing a coin: S = {Head, Tail} > Success: Head > Failure: Tail > Experiment: Rolling a dice: S = {1,2,3,4,5,6} > Success: Getting a six. > Failure: Getting any other number. > Experiment: Opinion polls: 5 = { Yes, No} » Success: Yes > Failure: No ‘As shown previously, in week 10, Bernoulli random variable > A random variable that takes on either the value 1 or 0 is called a Bernoulli random variable. » Let X be a Bernoulli random variable that takes on the value 1 with probability p. > The probability distribution of the random variable is > Expected value of a Bernoulli random varaible: E(X) =0x (1—p)+1xp=p > Variance of a Bernoulli random variable: Var(X) = p — p? = p(1— p), Note: Variance of a Bernoulli random variable is a quadratic p-p®. It is a downward facing parabola with a maximum point at x=-b/a which is % in this case. So, the largest variance in a Bernoulli trial is %.> The largest variance occurs when p= 3, when success and failure are equally likely. > In other words, the most uncertain Bernoulli trials, those with the largest variance, resemble tosses of a fair coin. Independent and Identically Distributed (IID) Bernoulli trials > A collection of Bernoulli trials defines iid Bernoulli random variables, one for each trial > The abbreviation iid stands for independent and identically distributed. Definition A collection of random variables is iid if the random variables are independent and share a common probability distribution IID Bernoulli trials are called Binomial distributions. For instance, when a coin is tossed three times: n=3 independent trials > Let m=3 independent Bernoulli trials. > Let p is probability of success. > The probabilities of outcomes of the independent trials are S.No | Outcome | Probabilities 1 [Gss)_[pxpxp 2 [Gsf) | pxpx(l-p) 3 |(sfis) [px 2) xp 4 [eff [px —p)x (1p) 5 |(fss) | (—e)xpxp 6 7 (ff) | (1=p) x px (l= p) px PX Pn=3 independent trials, X = number of successes > Let m= 3 independent Bernoulli trials. > Let p is probability of success. = number of successes in 3 independent trials. > The probabilities of outcomes of the independent trials are ‘SINo | Outcome | Number of | Probabilities 1 |Gsa)_|3 pxpxp 2 [sf) [2 px px (1—p) 3 [(sfs)_ [2 px(l—p) XP 4 | tsffy [a px (1~ p) x (1~p) 3 [(fss)__[2 G@=p)xpxp 6 (fst) [a (1=p) xp x (1—p) 7_[ffs)_ [1 =p) x (1 — a) xP Let n= 3 independent Bernoulli trials. Let p is probability of success. Let X = number of successes in 3 independent trials, The probability distribution of X vvvy x 0 1 2) 3 P(X =i) | C= oP [3x px (= py | 3x px (1p) | For n independent trials, n independent trials, X = number of successes > Consider any outcome that results in a total of / successes. > This outcome will have a total of j successes and (n ~ i) failures. > Probability off success and (ni) failures = p! x (1— p)(-9 > There number of different outcomes that result in i successes and (n— i) failures=(*) > The probability of i successes in n trials is given by = a) xp x (1p)? P(X=