ST2334 Midterms Cheatsheet v1.0 (2019-05-01) 7.

7. The Inclusion-Exclusion Principle • Each possible value x of X represents an event that is a • F(x) is a non-decreasing function: x1 < x2 ⇒ F(x1 ) ≤
by Julius Putra Tanu Setiaji, page 1 of 2 ∪
n ∑
n ∑ ∑
n−1 n subset of the sample space S F(x2 )
Pr( Ai ) = Pr(Ai ) − Pr(Ai ∩ Aj ) +• If S has elements that are themselves real numbers, we • 0 ≤ F(x) ≤ 1
i=1 i=1 i=1 j=i+1 take X(s) = s. In this case Rx = S
1 Basic Concepts of Probability 2.6 Mean and Variance of an R.V.
• Sample Space (S) = set of all possible outcomes of a sta- ∑ n−1
n−2 ∑ ∑n 2.2 Equivalent Events 2.6.1 Expected Value / Mean / Mathematical Expectation
Pr(Ai ∩ Aj ∩ Ak ) − ... 2.2.1 Definition ∑ ∑
tistical experiment i=1 j=i+1 k=j+1 • Discrete: E(X) = µX = i xi f (xi ) = x xf (x)
• Sample Points = An element of the sample space • Let E be an experiment in sample space S. Let X be an 1 1 ∑
8. If A ⊂ B, then Pr(A) ≤ Pr(B) R.V. defined on S, and Rx its range space, i.e. X : S → R • If f (x) = N for each of the∫N values of x, E(X) = N i xi
• Event = Subset of a sample space ∞
1.7 Conditional Probability, P (A | B) • Let B be an event w.r.t. RX , i.e. B ⊂ RX • Continuous: E(X) = µX = −∞ xf (x) dx
• Sample space = sure event, subset of S = ∅ = null event
• Mutually exclusive/disjoint if A ∩ B = ∅
Pr(A∩B)
• Pr(A | B) = Pr(B) , if Pr(A) , 0 • Suppose A = {s ∈ S | X(s) ∈ B} • Remark: The expected value exists provided the sum/in-
• Contained: A ⊂ B ≡ B ⊃ A. (A consists of all sample points s in S for which X(s) ∈ B) tegral exists
• For fixed A, Pr(B | A) satisfies the postulates of probabil-• A and B are equivalent events, and Pr(B) = Pr(A)
• If A ⊂ B and B ⊃ A, then A = B ity. 2.6.2 Expectation of a function of an R.V.
1.1 Basic Properties 2.2.2 Example ∀ g(X) with p.f. fX (x)
• False positive: Pr(+ | condition) ∑
• Consider tossing a coin twice, S = {HH, HT , T H, T T }
1. A ∩ A′ = ∅ 6. (A ∪ B)′ = A′ ∩ B′ 1.7.1 Multiplication rule • Discrete: E[g(X)] = x g(x)fX (x)
∫
• Let X be no of heads, then R = {0, 1, 2}
2. A ∩ ∅ = ∅ 7. A∪(B∩C) = (A∪B)∩(A∪C) • Pr(A ∩ B) = Pr(A) Pr(B | A) = Pr(B) Pr(A | B), providing • A = {HH} equiv B = {2}, AX = {HT , T H} equiv B = {1}, • Continuous: E[g(X)] = ∞ g(x)fX (x) dx
3. A ∪ A′ = S 8. A∩(B∪C) = (A∩B)∪(A∩C) Pr(A) > 0, Pr(B) > 0 1 1 2 2 −∞
4. (A′ )′ = A ′ A3 = {T T } equiv B3 = {0}, A4 = {HH, HT , T H} equiv • Provided the sum/integral exists.
9. A ∪ B = A ∪ (B ∩ A ) • Pr(A ∩ B ∩ C) = Pr(A) Pr(B | A) Pr(C | A ∩ B)
5. (A ∩ B)′ = A′ ∪ B′ B4 = {2, 1} 2.6.3 Variance (σX2 = V (X))
10. A = (A ∩ B) ∪ (A ∩ B′ ) • Pr(A1 ∩ ... ∩ An ) = Pr(A1 ) Pr(A2 | A1 ) Pr(A3 | A1 ∩ 2.3 Discrete Probability Distributions
1.2 De Morgan’s Law A2 )... Pr(An | A1 ∩ ... ∩ An−1 ) • g(x) = (x − µX )2 , Let X be an R.V. with p.f. f (x)
2.3.1 Discrete R.V.
∪
n ∩n ∩
n ∪
n 1.7.2 The Law of Total Probability Let X be an R.V. If RX is finite or countable infinite, X is • σX2 = V (X) = E[(X − µX )2 ]
1. ( Ar )′ = (Ar )′ 2. ( Ar ) ′ = (Ar )′ {∑
• Let A , A , ..., A be a partition of sample space S (mutu- discrete R.V. 2
2 ] = ∫ x (x − µX ) fX (x)
r=1 r=1 r=1 r=1 1 2 n if X is discrete
ally exclusive and exhaustive events s.t. A i ∩ A j = ∅ for 2.3.2 Probability Function (p.f.) or Probability Mass • E[(X−µ ) ∞
1.3 Counting Methods
n Function (p.m.f.)
X
−∞
(x − µ X )2 fX (X) dx if X is continuous
1.3.1 Multiplication & Addition Principle i , j and ∪i=1 Ai = S).
∑ ∑ • For a discrete R.V., each value X has a certain probability • V (X) ≥ 0, V (X) = E(X 2 ) − [E(X)]2
1.3.2 Permutation • Then Pr(B) = ni=1 Pr(B ∩ Ai ) = ni=1 Pr(Ai ) Pr(B | Ai ) √
• An arrangement of r objects from a set of n objects, r ≤ n, 1.7.3 Bayes’ Theorem f (x). Such a function f (x) is called the p.f. • Standard deviation = σX = V (X)
order taken into consideration. • The collection of pairs (xi , f (xi )) is prob distribution of X
• Let A1 , A2 , ..., An be a partition of S 2.6.4 K-th moment of X
n! • The probability of X = xi denoted by f (xi ) must satisfy:
• n distinct objects taken r at a time = n Pr = (n−r)! Pr(Ak ) Pr(B|Ak ) Pr(Ak ) Pr(B|Ak ) • Definition: E(X k ), use g(x) = xk in expectation of a fn
• Pr(Ak | B) = ∑n = , k ∈ [1, n] 1. f∑(xi ) ≥ 0 ∀ xi
• In a circle: (n − 1)! i=1 Pr(Ai ) Pr(B|Ai )
Pr(B)
2. ∞
2.6.5 Properties of Expectation
∑ i=1 f (xi ) = 1 1. E(aX + b) = aE(X) + b
• Not all are distinct: kr=1 nk = n, n Pn1 ,n2 ,...,nk = n !n n!!...n ! 1.8 Independent Events 2.4 Continuous Probability Distributions
1 2 k • Definition: iff Pr(A ∩ B) = Pr(A) Pr(B) 2. V (X) = E(X 2 ) − [E(X)]2
1.3.3 Combination 2.4.1 Continuous R.V. 2
1.8.1 Properties Suppose that RX is an interval or a collection of intervals, 3. V (aX + b) = a V (X)
• No of ways selecting r from n objects w/o regarding order
(n) n! • Suppose Pr(A) > 0, Pr(B) > 0, A and B are independent: then X is a continuous R.V. 2.7 Chebyshev’s Inequality
• r =n Cr = r!(n−r)! , n Cr × r! =n Pr – Pr(B | A) = Pr(B) and Pr(A | B) = Pr(A)
(n) r n−r
2.4.2 Probability Density Function (p.d.f.) • Let X be an R.V. with E(X) = µ, V (X) = σ 2
• r = binomial coefficient of the term a b in binomial – A and B cannot be mutually exclusive (and vice versa) • Let X be a continuous R.V.
• The sample space S and ∅ are independent of any event • p.d.f. f (x) is a function satisfying: • ∀ k > 0, Pr(|X − µ| > kσ ) ≤ 12
expansion of (a + b)n : k
( ) ( n ) • If A ⊂ B, then A and B are dependent unless B = S 1. f∫ (x) ≥ 0 ∀ x ∈ RX ∫ • Alternatively, Pr(|X − µ| ≤ kσ ) ≥ 1 − 12
1. nr = n−r for r = 0, 1, ..., n ∞ k
( ) ( ) (n−1) Warning: Indep events can’t be shown using Venn Diagram, 2. f (x) dx = 1 or −∞ f (x) dx = 1 as f (x) = 0 ∀ x < RX • Holds for all distributions with finite mean and variance
2. nr = n−1 r + r−1 for 1 ≤ r ≤ n hence calc! Cannot use intuition RX
(n) ∫d • Gives a lower bound but not exact probability.
3. r = 0 for r < 0 pr r > n 1.8.2 Theorem 3. ∀ c, d : c < d (i.e. (c, d) ⊂ RX ), Pr(c ≤ X ≤ d) = c f (x) dx 3 MA1521 Shit
1.4 Relative frequency (fA ) If A, B are indep, then so are A and B′ , A′ and B, A′ and B′ .
2.4.3 Remarks 3.1 Taylor Series of f at a
n 1.8.3 n Independent Events ∫d
fA = nA , event A in n repetitions of experiment E, nA = no ≤ ≤ ∑ f (k) (a)
of times that event A occured among the n repetitions.
• Pairwise Independent Events: • Pr(c X d) = c
f (x) dx represents area under the f (x) = ∞ k=0 k!
(x − a)k
Events A1 , A2 , ..., An are pairwise indep graph of the p.d.f. f (x) between x = c and x = d
1.4.1 Properties 3.2 MacLaurin Series
iff Pr(Ai ∩ Aj ) = Pr(Ai ) Pr(Aj ) • Let x0 be a fixed value, Pr(X = x0 ) = 0 ∑ f (n) (0) n
1. 0 ≤ fA ≤ 1 • Mutually Independent: • ≤ and < can be used interchangeably in a prob statement. Taylor series of f at 0, i.e. f (x) = ∞ n=0 n! x
2. fA = 1 iff A occurs every time among the n repetitions
Events A1 , A2 , ..., An are (mutually) independent iff for • Pr(A) = 0 does not necessarily imply A = ∅ 3.3 List∑ of common MacLaurin Series
3. fA = 0 off A never occurs among the n repetitions • RX ∈ [a, b] ⇒ f (x) = 0 ∀ x < [a, b] 1 = ∞ xn , −1 < x < 1, R = 1
any subset {Ai1 , Ai2 , ..., Aik } of A1 , A2 , ..., An , • 1−x
4. Events A and B are mutually exclusive → fA∪B = fA +fB ∑ n=0
1 = ∞ (−1)n xn , −1 < x < 1, R = 1
5. fA "stabilises" near some definite numerical value as the Pr(Ai1 ∩ Ai2 ∩ ... ∩ Aik ) = Pr(Ai1 ) Pr(Ai2 )... Pr(Aik )
2.5 Cumulative Distribution Function (c.d.f.) • 1+x
• Let X be an R.V., discrete or continuous. ∑n=0
1.8.4 Remarks 1 = ∞ x2 n, −1 < x < 1, R = 1
experiment is repeated more and more times. • F(x) is a c.d.f. of X where F(x) = Pr(X ≤ x) •
1.5 Axioms of Probability • A ,
1 2 A , ..., A n are mutually independent ⇔ for any pair of 1+x 2 n=0
events Aj , Ak where j , k, the multiplication rule holds, • F(x) = ∑
2.5.1 c.d.f. for Discrete ∑ R.V. ∑ (−1)n−1 xn
1. 0 ≤ Pr(A) ≤ 1 f (t) = Pr(X = t) • ln(1 + x) = ∞ n=1 n , −1 < x < 1, R = 1
t≤x t≤x
2. Pr(S) = 1 for any 3 distinct events, the multiplication rule holds, • c.d.f. of a discrete R.V. is a step function ∑∞ (−1)n x2n+1
3. If A1 , A2 , ... are mutually exclusive (disjoint),∑ and so on Pr(A1 ∩ A2 ∩ ... ∩ An ) = Pr(A1 ) Pr(A2 )... Pr(An ).• ∀ a, b s.t. a ≤ b, Pr(a ≤ X ≤ b) = Pr(X ≤ b) − Pr(X < a) = • sin x = n=0 (2n+1)! , −∞ < x < ∞, R = ∞
i.e. Ai ∩Aj = ∅ when i , j, then Pr(∪∞ ∞ In total there are 2n − n − 1 different cases.
i=1 Ai ) = i=1 Pr(Ai ) F(b) − F(a− ) where a− is the largest possible value of X • cos x = ∑∞ (−1)n x2 n , −∞ < x < ∞, R = ∞
In particular, if events A and B are mutually exclusive, • Mutually indep ⇒ pairwise indep (not the converse) that is strictly less than a n=1 (2n)!
• Suppose A1 , A2 , ..., An are mutually indep events, let Bi = ∑ xn , −∞ < x < ∞, R = ∞
then Pr(A ∪ B) = Pr(A) + Pr(B) ′ • RX ⊂ Z, a, b ∈ Z ⇒ • ex = ∞
1.6 Properties of Probability A i or A i , i ∈ [1, n]. Then B 1 , B 2 , ..., B n are also mutually – Pr(a ≤ X ≤ b) = Pr(X = a or a + 1 or ... or b) = F(b) − F(a −
n=0
∑
n!
n
• tan −1 x = ∞ (−1) x2n+1 , −1 ≤ x ≤ 1, R = 1
1. Pr(∅) = 0 indep events. 1) n=0 2n+1
∑
2. If A1 , A2 , ..., A∑n are mutually exclusive events, then 2 Concepts of Random Variables – Taking a = b, Pr(X = a) = F(a) − F(a − 1) • 1 = ∞ n=1 nx
n−1 , −1 < x < 1, R = 1
(1−x)2
Pr(∪ni=1 Ai ) = ni=1 Pr(Ai ) 2.1 Random Variable
∫∞
2.5.2 c.d.f. for Continuous R.V. ∑
2.1.1 Definition • 1 = 12 ∞ n=2 n(n − 1)x
n−2 , −1 < x < 1, R = 1
3. Pr(A′ ) = 1 − Pr(A) Let S be sample space assoc with experiment E. R.V. is a • F(x) = −∞ f (t) dt (1−x)3
4. Pr(A) = Pr(A ∩ B) + Pr(A ∩ B ) ′ ∑ ( )
function X, which assigns a number to every element s ∈ S • f (x) = dF(x) if the derivative exists • (1 + x)k = ∞ k n
n=0 n x , −1 < x < 1, R = 1
5. Pr(A ∪ B) = Pr(A) + Pr(B) − Pr(A ∩ B) 2.1.2 Notes dx
n(n−1) n(n−1)(n−2) 3
6. P r(A ∪ B ∪ C) = Pr(A) + Pr(B) + Pr(C) − Pr(A ∩ B) − Pr(B ∩ • X is a real-valued function • Pr(a ≤ X ≤ b) = Pr(a < X ≤ b) = F(b) − F(a) n
• (1 + x) = 1 + nx + 2! x2 + x + ..., −1 < x <
3!
C) − Pr(A ∩ C) + Pr(A ∩ B ∩ C) • Range space of X, R X = {x | x = X(s), s ∈ S}. 1, R = 1
 4 7
ST2334 Midterms Cheatsheet v1.0 (2019-05-01) =√ x (4x3 ) = √ 4x
by Julius Putra Tanu Setiaji, page 2 of 2 (x4 )3 +2 x1 2+2
∫a ∫x
d
3. dx d
f (t)dt = − dx a f (t)dt
x
3.4 Indefinite ∫ Integral ∫ x4 ∫ 4 ∫ 2
4. dx d f (t)dt = dxd x f (t)dt − d x f (t)dt
Denoted by f (x)dx = F(x) + C x2 a dx a
3.5 ∫ Rules of Indefinite Integration 3.10 Integration Methods
∫ • Integration by
1. kf (x)dx = k f (x)dx ∫ Substitution :
∫ ∫ Use the form f (g(x))dg(x) OR use a dummy variable to
2. −f (x)dx = − f (x)dx
∫ ∫ ∫ get to a form in the Integral Formulae (taking into account
3. [f (x) ± g(x)]dx = f (x)dx ± g(x)dx chain rule)
3.6 Integral Formulae Integral Sub Use identity
Function
∫ Integral a2 − u 2 u = a sin θ 1 − sin2 θ = cos2 θ
∫ cot xdx ln(sin x) + C a2 + u 2 u = a tan θ 1 + tan2 θ = sec2 θ
∫ sec x tan xdx sec x + C u 2 − a2 u = a sec θ sec2 θ − 1 = tan2 θ
∫ x cot xdx
csc csc x + C •∫ Integration by ∫ Part :
sec 2 xdx tan x+C uv ′ dx = uv − u ′ vdx
∫
csc2 xdx − cot x + C Choose u by LIATE (Logarithmic, Inverse trigo, Algebraic,
∫ Trigo, Exponential)
xn dx xn+1 + C, n , −1, n rational
∫ n+1 3.11 Derivative Formulae
√ 1 dx sin−1 ( xa ) + C Function Derivative
∫ a 2 −x 2
(f (x))n nf ′ (x)f (x)n−1
1 dx 1 tan−1 ( x ) + C
∫ a2 +x2∫ a a sin f (x) f ′ (x) cos f (x)
1dx
∫ = dx x+C cos f (x) −f ′ (x) sin f (x)
ex dx ex + C tan f (x) f ′ (x) sec2 f (x)
∫
x ax cot f (x) −f ′ (x) csc2 f (x)
∫ a dx ln a
sec f (x) f ′ (x) sec f (x) tan f (x)
ln xdx x ln x − x + C
∫
1 dx csc f (x) −f ′ (x) csc f (x) cot f (x)
ln x + C
∫ x af (x) f ′ (x)af (x) ln a
sin kxdx − coskkx + C k 0
∫
cos kxdx sin kx + C ef (x) f ′ (x)ef (x)
∫ k
f ′ (x)
2 tan x − x + C
∫tan xdx loga f (x) f (x) ln a
∫ sec xdx ln(sec x + tan x) + C f ′ (x)
ln f (x)
csc xdx ln(csc x − cot x) + C f (x)
f ′ (x)
3.7 Riemann (Definite) Integrals sin−1 f (x) √
∑ 1−f (x)2
Riemann sum on f on [a, b] ≈ nk=1 f (ck )∆x
∑n f ′ (x)
Exact area = lim k=1 f (ck )∆x
n→∞ cos−1 f (x) −√
Riemann Integral of f over [a, b]: 1−f (x)2
∫b ∑ f ′ (x)
f (x)dx = lim nk=1 f (ck )∆x tan−1 f (x)
a n→∞ 1+f (x)2
3.8 Rules of Definite Integrals 3.12 Rules of Differentiation
∫a ∫b ∫b • (kf )′ (x) = kf ′ (x)
1. f (x)dx = 0, kf (x)dx = k f (x)dx
∫ba ∫aa a • (f ± g)′ (x) = f ′ (x) ± g ′ (x)
2. a f (x)dx = − b f (x)dx • dxd uv = u dv + v du
∫b ∫b ∫b ( )′ dx dx
3. a [f (x) ± g(x)] = a f (x) ± a g(x) f f ′ (x)g(x)−f (x)g ′ (x)
• g (x) =
∫b ∫b (g(x))2
4. If f (x) ≥ g(x) on [a, b], then a f (x)dc ≥ a g(x)dx d f (g(x)) = f ′ (g(x))g ′ (x) or dy = dy · du
∫b • dx dx du dx
If f (x) ≥ 0 on [a, b], then a f (x)dx ≥ 0
5. If f is continuous on the interval joining a, b and c,
∫b ∫c ∫c
then a f (x)dx + b f (x)dx = a f (x)dx
3.9 Fundamental Thm of Calculus
F ′ (x) = f (x) If F is an antiderivative of f on [a, b], then
∫b ∫b
a
F ′ (x)dx = a f (x)dx = F(b) − F(a)
∫ f be continuous on [a, b]. Then
x‘ Let
d x f (t)dt = f (x)
dx a ∫
Note the 2 x’s: on dx d and x and f (t) is indep of x
∫2 ∫ x a√ √
1. dx d t 2 dt = 0, dxd sin tdt = sin x
(0 ) 0 ( )
∫ x4 ∫ x4 4
2. dx d √ t dt = d √ t dt dx
1 3 dx 4 1 3 dx
t +2 t +2