ENGINEERING DATA ANALYSIS 1 – PROBABILITY

.Fundamental Principle of Coun�ng.

Individual event/object w/iden�cal elements, all are taken (𝒏𝒏 = 𝒓𝒓)
let: m # 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒/𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
n!
n # 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒/𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 N=
p! q! s! …
N = 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤
and 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎
N=m × n =
(𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒)!

or/either
N=m+ n Cyclic Permuta�on
i. Both direc�on is not the same

Group, subgroup event/object N = (n − 1)!

let: m # 𝑜𝑜𝑜𝑜 𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔, 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
ii. Both direc�on is the same
n # 𝑤𝑤𝑤𝑤𝑤𝑤ℎ𝑖𝑖𝑖𝑖 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
N = 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤  Ex. Wearables

N = m! × n! 1
N = (n − 1)!
2

.PERMUTATION.
 Ordered Arrangement of “𝑛𝑛” element
taken "𝑟𝑟” at a �me.
 If “𝑟𝑟” is the taken arrangement, .COMBINATION.
then (𝑛𝑛 − 𝑟𝑟) is the not taken arrangement.
 Grouping arrangement of elements
regardless of order.
n! 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 Combina�on of “𝑛𝑛” dis�nct objects
N = n Pr = =
(n − r)! (𝑛𝑛𝑛𝑛𝑛𝑛 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎)! taken "𝑟𝑟” at a �me.

n! 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎
All are taken (𝒏𝒏 = 𝒓𝒓) N = n Cr = =
(n − r)! × r! (𝑛𝑛𝑛𝑛𝑛𝑛 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎)!
N = n Pn = 𝑛𝑛!
n Pr
Box Method (Ex. 𝑛𝑛 = 4) n Cr = → C = P/r!
r!

N = 4 × 3 × 2 × 1 = 𝑛𝑛!

w/iden�cal elements, taken "𝒓𝒓” at a �me.

n!
N=
(n − r)! p! q! s! …

𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎
=
(𝑛𝑛𝑛𝑛𝑛𝑛 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎)! (𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒)!
 ENGINEERING DATA ANALYSIS 1 – PROBABILITY
.BASIC PROBABILITY.
PROBABILITY RULES
Probability of an Event (PE) Rule of Product
# 𝑠𝑠𝑠𝑠𝑠𝑠𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑓𝑓𝑓𝑓𝑓𝑓 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 i. Independent
PE =
# 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
Probability of event A and B
𝐷𝐷𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
PE =
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤
PA PB
� 𝑛𝑛 Cr �� 𝑛𝑛2 Cr2 �� 𝑛𝑛3 Cr3 � …
PE = 1 1 → 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶
𝑛𝑛𝑇𝑇 Cr𝑇𝑇

P(A ∩ B) = P(A and B) = P(A) × P(B)

Maximum Probability of an Event (𝐏𝐏𝐦𝐦𝐦𝐦𝐦𝐦 )
P(A𝐵𝐵� ) = P(A not B) = P(A) − P(A) × P(B)
# 𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = # 𝑜𝑜𝑜𝑜 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
P=1 ii. Dependent
 One event has an effect on the outcome of the next
event.
Probability that is not Event (𝐏𝐏𝐄𝐄� )
Probability of event A and B,
𝐏𝐏𝐄𝐄� = 1 − PE Given that A has occurred:

Condi�onal Probability (Baye’s Theorem) P(A and B) = P(A) × P(B|A)

 Probability of an event occurring given that another event has
occurred.

“Probability of event B given that event A has occurred.” Rule of Sum

P(A ∩ B) i. Mutually Exclusive
P(A|B) =
P(B)
Probability of event A or B
condi�on event

PA PB
Mathema�cal Expecta�on
𝑛𝑛
+E → 𝑤𝑤𝑤𝑤𝑤𝑤
E = � Pn ∙ 𝑥𝑥𝑛𝑛
−E → 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙
𝑛𝑛=0 P(A ∪ B) = P(A or B) = P(A) + P(B)
𝑤𝑤ℎ𝑒𝑒𝑒𝑒𝑒𝑒:
𝑥𝑥𝑛𝑛 = 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
Pn = 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 ii. Non – Mutually Exclusive
Probability of event A or B
ODDS (chance)
Odds for an event to occur: PA PB
P(E)
P=
1 − P(E)

Odds against an event to occur: P(A ∪ B) = P(A) + P(B) − P(A) × P(B)

1 − P(E)
P=
P(E)
�B
P(A �) = 1 − P(A ∪ B)
 ENGINEERING DATA ANALYSIS 1 – PROBABILITY
.Probability Distribu�on Func�on.
Discrete Probability Distribu�on
Binomial Distribu�on or Repeated Trials Mul�nomial Distribu�on
 Use when given a probability of success.  generaliza�on of the binomial distribu�on.

Condi�ons:
n!
i. Only 2 possible outcomes. 𝑃𝑃 = ∙ (𝑃𝑃1 )𝑥𝑥1 (𝑃𝑃2 )𝑥𝑥2 (𝑃𝑃3 )𝑥𝑥3 …
ii. Probability of each outcome must be 𝑥𝑥1 ! ∙ 𝑥𝑥2 ! ∙ 𝑥𝑥3 ! …
independent and constant.

𝑤𝑤ℎ𝑒𝑒𝑒𝑒𝑒𝑒:
P = n Cr ⋅ pr ∙ q(n−r)
𝑃𝑃 = 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃. 𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
𝑥𝑥 = # 𝑖𝑖𝑖𝑖 𝑜𝑜𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
where: n = num. of trials 𝑛𝑛 = 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 # 𝑖𝑖𝑖𝑖 𝑎𝑎𝑎𝑎𝑎𝑎 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
r = num. of successful trials
p = P of success Hyper – Geometric Distribu�on
 describes the probability of successes in draws,
q = P of failure
 without replacement, from a finite popula�on of size that
contains exactly objects with that feature,
 each draw is either a success or a failure.

( P CS )((N−P) C(N−S) )
P=
Poisson Distribu�on N Cn
 Approxima�on/ limi�ng case of Binomial Distribu�on, 𝑤𝑤ℎ𝑒𝑒𝑒𝑒𝑒𝑒:
when number of trials (n) gets very large and the probability of 𝑃𝑃 = 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆
success (P) is very small. 𝑆𝑆 = 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆
 Keyword: Average 𝑁𝑁 = 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
𝑛𝑛 = 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
Condi�ons:
i. Any # of possible outcomes during a �me period.
ii. Probability of each outcome must be Con�nuous Probability Distribu�on(shared w/ EDA 2)
independent and constant.
iii. The probability of an event is propor�onal Probability Density Func�on (PDF)
to the length of the �me period.
𝑓𝑓(𝑥𝑥) ≥ 0
λx ∙ e−λ 𝑃𝑃𝑃𝑃𝑃𝑃 �
P= 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 = 1
x!
𝑏𝑏
σ2 = npq where: 𝑃𝑃(𝑥𝑥) = � 𝑓𝑓(𝑥𝑥) = 1
𝑎𝑎
n = num. of trials
𝑥𝑥 = num. of successful trials
• Normal
p = P of success 1 𝑥𝑥 2
F(𝑥𝑥 ) = e− 2
q = P of failure √2𝜋𝜋
2
σ = 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣 • Exponen�al
e−x , 𝑥𝑥 ≥ 0
F(𝑥𝑥 ) = �
𝜆𝜆 = np = 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 0 , 𝑥𝑥 < 0
≤ 0.05 , 𝑛𝑛≥20 • Uniform
𝑝𝑝 ≤ 0.1 , 𝑛𝑛≥100 1 , 0 ≤ 𝑥𝑥 ≤ 1
F(𝑥𝑥 ) = �
0 , 𝑜𝑜𝑜𝑜ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
• Gamma
• Beta

Cumula�ve Distribu�on Func�on

 Complementary, Folded, Inverse, Empirical

CDF = � PDF