Moment generating function (MGF) — definition, use

4.9.5 · Maths › Probability Theory & Statistics

$e^{tX}$ moments kyun generate karta hai?

Trick hai $e^{tX}$ ki Taylor series. Ise scratch se derive karo:

$e^{tX} = \sum_{n = 0}^{\infty} \frac{( tX ) ^{n}}{n !} = 1 + tX + \frac{t ^{2} X ^{2}}{2 !} + \frac{t ^{3} X ^{3}}{3 !} + \dots$

Dono taraf expectation lo ( $E$ ki linearity se):

$M_{X} (t) = E [e^{tX}] = 1 + tE [X] + \frac{t ^{2}}{2 !} E [X^{2}] + \frac{t ^{3}}{3 !} E [X^{3}] + \dots$

$n$ -th moment kaise nikaalein

Toh:

$M_{X} (0) = 1$ hamesha (ye hai $E [e^{0}] = E [1]$ ). Hamesha pehle yahi check karo.
$M_{X}^{'} (0) = E [X]$ (mean).
$M_{X}^{''} (0) = E [X^{2}]$ , isliye $Var (X) = M_{X}^{''} (0) - (M_{X}^{'} (0))^{2}$ .

Figure — Moment generating function (MGF) — definition, use

Key PROPERTIES (aur ye kyun hold karte hain)

Common Mistakes (Steel-manned)

MGF kab USE karna hai (80/20)

20% jo 80% value deta hai:

Mean/variance compute karo bina integrals ghiste hue — bas differentiate karo.
Ek distribution identify karo uska MGF pehchaan ke (uniqueness).
Independents ke sum ki distribution nikalo (MGFs multiply karo).
Limit theorems prove karo (CLT proof: dikhao ki MGFs $e^{t^{2} /2}$ par converge karte hain).

Recall Feynman: ek 12-saal ke bachche ko explain karo

Socho $X$ ek mystery shape hai jo clay se bani hai. Ise describe karne ke liye tum facts list karte: uska middle kahan hai, kitna wide hai, kitna tedha — ye moments hain. Inhe ek ek karke likhna slow hai. MGF ek magic box hai: tum ise ek dial number $t$ dete ho, aur ye secretly SAARE facts ek saath store kar leta hai, folded together. Fact number $n$ nikaalte ke liye, tum "dial $n$ baar hilate ho" ( $n$ derivatives lete ho) aur phir dial wapas zero par set karte ho. Exactly woh ek fact bahar aa jaata hai, clean. Bonus magic: agar tum do independent shapes ko jod do, toh jodi hui shape ka box sirf do boxes multiplied hai — isliye mathematicians isse pyaar karte hain.

Flashcards

Definition of the MGF of

X

M_{X} (t) = E [e^{tX}]

t = 0

ke aaspaas ek interval mein finite.

How do you get the

n

-th moment from the MGF?

E [X^{n}] = M_{X}^{(n)} (0)

n

-th derivative

t = 0

par evaluate ki gayi.

Why does

e^{tX}

generate moments?

Its Taylor series

\sum \frac{t ^{n} X ^{n}}{n !}

hai;

E

lene se

t^{n}

ka coefficient

E [X^{n}] / n!

ban jaata hai.

What is

M_{X} (0)

for every random variable?

1

, kyunki

E [e^{0}] = E [1] = 1

— ise sanity check ki tarah use karo.

MGF of

Y = a X + b

M_{Y} (t) = e^{b t} M_{X} (a t)

MGF of

X + Y

when

X ⊥ Y

M_{X + Y} (t) = M_{X} (t) M_{Y} (t)

(multiplication, independence chahiye).

Variance in terms of the MGF?

Var (X) = M_{X}^{''} (0) - (M_{X}^{'} (0))^{2}

MGF of Exponential

(λ)

and its domain?

M_{X} (t) = \frac{λ}{λ - t}

for

t < λ

MGF of Normal

(μ, σ^{2})

M_{X} (t) = exp (μ t + \frac{1}{2} σ^{2} t^{2})

MGF of Binomial

(n, p)

and why?

(1 - p + p e^{t})^{n}

, kyunki ye

n

independent Bernoulli MGFs ka product hai jo multiply kiye gaye.

Why might an MGF not exist?

Agar

E [e^{tX}] = \infty

0

ke paas ho (heavy tails, jaise Cauchy); tab characteristic function use karo.

What does uniqueness of MGF give you?

Agar do MGFs

0

ke aaspaas ek interval par agree karte hain, toh dono distributions identical hain.

Connections

Probability Distributions — har named distribution ka ek signature MGF hota hai.
Expectation and Variance — MGF derivatives inhe directly recover karte hain.
Characteristic Function — $E [e^{i tX}]$ wala cousin jo hamesha exist karta hai.
Central Limit Theorem — MGFs ke $e^{t^{2} /2}$ par convergence se prove hota hai.
Independence (Probability) — multiplication rule ke peeche yahi condition hai.
Taylor Series — wo engine jo moments ko "bahar girne" deta hai.
Cumulant Generating Function — $ln M_{X} (t)$ , cumulants generate karta hai.

Moment generating function (MGF) — definition, use

YE HAI KYA?

$e^{tX}$ moments kyun generate karta hai?

$n$ -th moment kaise nikaalein

Key PROPERTIES (aur ye kyun hold karte hain)

Worked Examples

Common Mistakes (Steel-manned)

MGF kab USE karna hai (80/20)

Flashcards

Connections

Concept Map

Deep Dive