Scribd
Upload
20 views20 pages

Lecture Ba 03

This document covers key concepts in statistics relevant to business analytics, including geometric series, combinatorics, random variables, expectations, and standard deviation. It explains the sum of geometric series, combinations and permutations, and introduces probability distribution functions (PDF) and cumulative distribution functions (CDF). Additionally, it discusses variance, covariance, correlation, and various distributions such as normal and binomial distributions.

Uploaded by

Nathalie He
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
20 views20 pages

Lecture Ba 03

This document covers key concepts in statistics relevant to business analytics, including geometric series, combinatorics, random variables, expectations, and standard deviation. It explains the sum of geometric series, combinations and permutations, and introduces probability distribution functions (PDF) and cumulative distribution functions (CDF). Additionally, it discusses variance, covariance, correlation, and various distributions such as normal and binomial distributions.

Uploaded by

Nathalie He
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 20

MATHEMATICS & STATISTICS FOR BUSINESS

ANALYTICS
LECTURE #3
STATISTICS

 Sum of geometric series and combinatorics

 Random variables
 Expectations
 Standard deviation and variance
 Parametric distributions
SUM OF GEOMETRIC SERIES

Infinite Series Example


𝑆 = 1 + 𝛿 + 𝛿2 + 𝛿3 + ⋯ 𝛿 <1 𝛿 = 0.50
1 1
𝑆= = =2
1 − 𝛿 0.50
𝛿𝑆 = 𝛿 + 𝛿 2 + 𝛿 3 + 𝛿 4 … 𝑆 − 1 = 𝛿 + 𝛿2 + 𝛿3 + 𝛿4 + ⋯

Be Careful
𝛿𝑆 = 𝑆 − 1 𝑆 = 𝛿 + 𝛿2 + 𝛿3 + ⋯
1=𝑆 1−𝛿
1 𝛿
𝑆= 𝑆=
1−𝛿 1−𝛿
SUM OF GEOMETRIC SERIES

Finite Series Example


𝑆 = 1 + 𝛿 + 𝛿 2 + ⋯ + 𝛿 𝑛−1 𝛿 = 0.50, 𝑛 = 3
1 − 𝛿𝑛
𝑆=
1−𝛿
𝛿𝑆 = 𝛿 + 𝛿 2 + 𝛿 3 + ⋯ + 𝛿 𝑛 𝑆 − 1 + 𝛿𝑛 = 𝛿 + 𝛿2 + 𝛿3 + ⋯ + 𝛿𝑛 1 − 0.503
=
0.50
= 2 − 0.502
𝛿𝑆 = 𝑆 − 1 + 𝛿 𝑛
8 1
𝑛
1−𝛿 =𝑆 1−𝛿 = −
4 4
1 − 𝛿𝑛 3
𝑆= =1
1−𝛿 4
COMBINATORICS

Combination
 How many different groups of k items can you create from n items?
 Order does not matter

Example In general
Given {a, b, c}, you can create 3 combinations of 2 objects: 𝑛!
𝑛
𝐶𝑘𝑛 = =
 {a, b} 𝑘 𝑘! 𝑛 − 𝑘 !
 {a, c}
 {b, c}
COMBINATORICS

Permutation
 How many different groups of k items can you create from n items?
 Order does matter

Example
In general
Given {a, b, c}, you can create 6 permutations of 2 objects:
 There are k! ways to order k objects
 {a, b}  {b, a}
𝑛!
 {a, c}  {c, a} 𝑃𝑘𝑛 = 𝑘! × 𝐶𝑘𝑛 =
𝑛−𝑘 !
 {b, c}  {c, b}
MINI-QUIZ #1
RANDOM VARIABLES

 Random variable
 More than one possible outcome
 Some outcomes may be more likely
 Probability of all outcomes must be 100%

 Discrete vs Continuous
 Discrete: P[X = 5]
 Continuous: P[0 ≤ Y ≤ 10]
RANDOM VARIABLES

Joint distribution
F↑ F↓  Conditional probability
M↑ 40% 10% P[𝑋|𝑌]
M↓ 20% 30%  Independent random variables
P 𝑋&𝑌 = P 𝑋 P[𝑌]  independence

Joint distribution + marginal distributions


F↑ F↓
M↑ 40% 10% 50%
M↓ 20% 30% 50%
60% 40%
RANDOM VARIABLE
The PDF and CDF are interrelated concepts in probability
theory. The PDF gives the probability of a continuous random
 Probability Distribution Function (PDF) variable taking on a specific value. At the same time, the CDF
 Discrete provides the cumulative probability of the random variable
being less than or equal to a given value.
 example: P[X = 1] = 20%, P[X = 2] = 60%, P[X = 3] = 20%
 σ𝑛𝑖=1 𝑝𝑖 = 1

 Continuous
10
 example: P[2 ≤ X ≤ 10] = ‫׬‬2 𝑝 𝑥 𝑑𝑥
∞
 ‫׬‬−∞ 𝑝 𝑥 𝑑𝑥 = 1

 Cumulative Distribution Function (CDF)


𝑎
 𝐹 𝑎 = 𝑃 𝑋 ≤ 𝑎 = ‫׬‬−∞ 𝑝 𝑥 𝑑𝑥
𝑑𝐹(𝑥)
 𝑝 𝑥 =
𝑑𝑥
MINI-QUIZ #2
EXPECTATIONS

 Expected value of X = E[X]


+∞
 Given PDF, 𝐸 𝑋 = σ𝑛
𝑖=1 𝑝𝑖 𝑥𝑖 = ‫׬‬−∞ 𝑝 𝑥 𝑥𝑑𝑥

 For random variables X and Y and constants a and b:


 E[aX + bY] = aE[X] + bE[Y]
 E[XY] ≠ E[X]E[Y] (unless X and Y are independent)
EXPECTATIONS

Conditional Expectations

 If X = 5 + 10Y then
 E[X] = 5 + 10E[Y]
 E[X|Y] = 5 + 10Y (treat Y as if it were known, as if it was constant)
 Independence ⇒ E[X|Y] = E[X]
PARAMETER ESTIMATES

 Parameter estimates vs their true values (𝜇ො vs 𝜇)

 Population vs sample
MINI-QUIZ #3
STANDARD DEVIATION & VARIANCE

 Variance
 Expectations 𝜎2 = 𝐸 𝑋 − 𝜇 2
= 𝐸 𝑋 2 − 𝐸[𝑋]2
1
 Population 𝜎 2 = σ𝑛𝑖=1 𝑥𝑖 − 𝜇 2
𝑛
1  Standard deviation
 Sample 𝜎෢2 = σ𝑛𝑖=1 𝑥𝑖 − 𝜇Ƹ 2
𝑛−1
+∞  𝜎 = 𝜎2
2 2
 Continuous 𝜎 = ‫׬‬−∞ 𝑝 𝑥 𝑥 − 𝜇 𝑑𝑥
 𝜎𝑎+𝑏𝑥 = |𝑏|𝜎𝑥
 a.k.a. volatility
COVARIANCE & CORRELATION

 Cov[X, Y]
 σX,Y = E[(X − μX)(Y − μY)]

(++) (−−) (++) (−−)


1 𝑛
 σෞ
X,Y = 𝑛−1 σ𝑖=1 𝑥𝑖 − 𝜇𝑋 𝑦𝑖 − 𝜇𝑌

𝜎𝑋,𝑌
 Corr[X, Y] = ρXY =

(++) (−−) (+−) (−+)


𝜎𝑋 𝜎𝑌

2
 𝜎𝑋+𝑌 = 𝜎𝑋2 + 𝜎𝑌2 + 2𝜌𝜎𝑋 𝜎𝑌

(+−) (−+) (+−) (−+)


𝑛
 if:
𝑌 = ෍ 𝑤𝑖 𝑋𝑖
𝑖
𝑛 𝑛
then:
𝜎𝑌2 = ෍ ෍ 𝑤𝑖 𝑤𝑗 𝜎𝑖,𝑗 = 𝐰′𝚺𝐰
𝑖 𝑗

𝑤1 𝜎1,1 ⋯ 𝜎1,𝑛
𝐰= ⋮ 𝚺= ⋮ ⋱ ⋮
𝑤𝑛 𝜎𝑛,1 ⋯ 𝜎𝑛,𝑛
MINI-QUIZ #4
NORMAL DISTRIBUTION

 a.k.a. Gaussian distribution


1 𝑥−𝜇 2
1 −2 𝜎
 PDF: 𝑒
𝜎 2𝜋

 𝑋~𝑁 𝜇, 𝜎 2
 Linear combination of independent normal variables is also normal
 Excel: Norm.Dist(…) and Norm.Inv(…)
OTHER DISTRIBUTIONS

 Bernoulli
 Beta
 Binomial
 Chi-squared
 Exponential
 F-distribution
 Poisson
 Student’s t
 Uniform

You might also like