This** preview**
has intentionally

**sections.**

*blurred***to view the full version.**

*Sign up*
This** preview**
has intentionally

**sections.**

*blurred***to view the full version.**

*Sign up*
**Unformatted text preview: **IME602: Probability and Statistics
Quiz-4, October 27, 2016 An A c0 t‘< Duration: 1 hour Maximum Marks: 10 Name: ....................................... Roll No: This is a closed book/note examination. Useful information, whenever necessary, is provided
along with the question. Attempt all questions. 1. Let X1 and X 2 denote your quiz-1 and 2 marks. Let us assume X1 and X 2 to be independent
and uniformly distributed in [0,10]. Let Y = X1 + X 2 denote the total marks in quiz-1 and 2.
Derive the density function of Y. Represent fy (y) pictorially. [Marks: 3] Info: If X1 and X2 are independent, fx1+x2 (y) = L: fx1 (x) sz (y — x)dx. 2. Let C denote the melting point of a newly developed material. Measurements of c, due to
error, are of the form: Mi = c + 61-, where ei’s are the errors. Let us take ei’s to be iid random
variables with mean 0 and variance 3. Then Mi’s are iid random variables too. We consider IV! = Z?=1Mi/n as the measurement for c. Using Chebyshev’s inequality, in quiz-3, we got
11 2 120 to be 90% sure that [V] is in c i 0.5. Using the central limit theorem, ﬁnd the new
minimum of n to ensure the same accuracy. [Marks: 2] Note: According to the central limit theorem, X = Z?=1Xi/n, where Xi’s are iid random
variables with mean ,u and variance 02, approximately follows N (u, 02 / n) for large n. 3. We want to estimate the distribution function value of an arbitrary random variable at an
arbitrary point, i.e., we want to estimate 9 = Fx(a) for any X and a. We have a random
sample of X, denoted by (X1,X2, , Xn). Then it makes sense to estimate 0 by the proportion of Xi’s that do not exceed a, i.e., 0 = (221:1 1x,-5a)/na where 1Xi5a = 1 if X,- S a, otherwise 1X,-sa = 0. Study the properties of g (as an estimator of 0). [Marks: 2.5] Hint: 1xi5a’5 are iid random variables. Identify their mean and variance ﬁrst. 4. Estimate k and p for N e gB in(k, p) using the method of moments. [Marksz 2.5] Hint: Geometric(p) has mean 1 / p and variance (1 — p) / p2. I am talking about geometric
distribution when the question is about negative-binomial distribution! i~ Km u(o)1o) (W3 X2” 0 (0,10) m imﬂdmAeW. “:2 w x1. (mama, Y Wm mm 'm wan. HM, #chzo A 4 ¢Y0'“°i‘-““@ Pom % € [0,?»011 by 41M) comramkizm {lommﬂm N063 «Mm =r{f OS) 2 "1% {k 1 e ‘0’”) Q owotmi’it HI
1,;
1
2:;
m
x
A
L_C \M‘m (gum) — WWW (0’3"!) ‘50 W. = VA“; M awm W
(30%me 16% 3600.10] Q3? H gem; One (AM (imam ‘HAK 5W1 WWW {Mica Fa“ Ward/3b 8M“ \02 <1
Tka in? 04m 710%»,qu
tvm 647 \o'na (xx/3 AM OWWMDUt V3 COMQ} \ x ‘
swam} 54%“ \pe mwﬁw in mm empty. WW § 0 ID m»
“‘“k¢N}—-~“" 2. MMJz; am; m (Mg): 76» TAU“ AV FILL C(MJTM’I Aim/f aﬁﬁﬁcximﬁda y‘allcN/s 3 ”V4“ 411 2'. _~ ‘ EMK
M... :2. ‘
Ilmdrtw, 7%" Qﬁuﬂ’é n“ h ”<413/4)‘ ""“""‘””—‘ x [email protected]
wﬂmm‘P(C—0§§R<c+05)>/09 __________________ @
E ‘P *0‘6’< F4~c£ o-s >>0
< 3A 4?; I22 , 9
W
1660.1)
:: 0'5 "05‘ > 0.6) ‘‘‘‘‘‘‘‘‘‘ KZ‘
" “mywm ’ C?
r 0; N
: l~2<f>(~ M3205 By W
a r
- ¢(——J:/;> 5°'05%45<-16453(‘§“o~n~(¢m¢~“ @
.a ()‘S‘ ‘
: - W7: $ ‘1‘64S ms 69' [,3 Mum-H3 MUM/32m 1" K.
\4
i J 1'6“
W3 E W :> “E 324% X; has :5me Al‘m‘m'LrA/ydh M ”MAM/$410. Av», PM“ ~ 2? EMMA]: 5%” 6M We Mag {0 Mug; ZSHOygﬁ‘ZH-ﬁ»; 772 on «we 4471M ﬂzaﬁarl—I'w. x.“ W RA mmgm“ Vomubw), ._ in“ )/H wrmqw c9: ting,
(-1 ’0 >=I~ﬁ<w {g V[&(1X£5a)2 ECﬁO-gl—[iQ/D} #5 M m &¢;mﬂ;»rc #8. Wt 64/»? demM/JL 2,5 414 ”SE (:0 ﬁlm” 012 Inhale. C) do‘nbim‘mca .‘ 537162 m ® => E‘*1:'bk,EYY|1: V10. » —» ' , ~
\mcx): “$53100: ko-bﬁ/bg‘; , , — -- ‘ * “ @ To mmm Laws, ’9 04; ngLEvaMB ‘03 4m wamﬂ ﬁ wowmél {mm MW‘. 3;: 1/; - ~~»@ Since 4kg «bow WVJHO'M Mm Lo Ml- M‘WUW} XE MHz for! '2 mg I; we efwr-fka w {2 - (Ts/“1+ EC‘h Wok: LB ane (McuW’m Mum th Mtr‘tuncz if NC; 3??? (Ida) Qc‘wdlgg, {MI Mm)
La. 1 MK/M be ﬁivw. ...

View
Full Document

- Fall '17
- Faiz Hamid
- Probability theory, 5%, 5°, Mama, Ilmdrtw