Prof: Norman Schürho¤ TAs: Zhihua Chen (Cissy), Natalia Guseva
1. University is accepting freshmen. This University can not accommodate
more than 1060 students. Assume that each applicant accepts with prob- ability 0.6. If University accepts 1700 students, what is the probability that it will have too many acceptance. 2. Random variable X is said to have Pareto distribution with parameter ; if it is continuously distributed with density: ( +1) x ; if x > 1 fX (xj ) = : 0; otherwise There is random sample x1 :::xn from P areto( ) available. Derive ML estimator b, prove its consistency and …nd its asymptotic dis- tribution. 3. Let X1 ; :::; Xn be a sample from a normal distribution with mean 1 and variance 1, let Y1 ; :::; Yn be a sample from a distribution with mean 2 and variance 2 and let Z1 ; :::; Zn be a sample from a distribution with mean 1 + 2 and variance 4. All three samples are independent. (a). Find the maximum likelihood estimators of 1 and 2 : (b). Find the Fisher Information Matrix (when n = 1). Use it to …nd the asymptotic distribution of the MLE’s as n ! 1: 4. (a). Let be a …xed, integrable random variable and Fn an arbitrary …ltration. Show that Xn = E( =Fn ) is a martingale. (b). Let X1 ; :::; Xn be a sequence of independent r.v.’s with mean 0. Set Fn = (X1 ; :::; Xn ) the -…eld generated by X1 ; :::; Xn : Let Mn = P n Xi : Show that Mn is a martingale. Now, take Mn = Sn2 n; where i=1 P n Sn = i=1 Xi and assume that the Xi all have variance 1. Show that Mn is a martingale. (c). Suppose you start with a franc and you are tossing a fair coin inde- pendently. If it turns up head you double your fortune, tails you go broke. Let Mn be your fortune at time n; i.e., if you let X1 ; :::; Xn be independent r.v.’s that are equal to 2 with probability 0.5 and to 0 with probability 0.5, then show Mn = X1 ::: Xn is a martingale.