0% found this document useful (0 votes)
3 views15 pages

Statistical Estimation and MLE Analysis

The document discusses statistical estimation, focusing on Maximum Likelihood Estimation (MLE) and its application to various distributions such as Bernoulli, Poisson, Exponential, and Normal. It also covers statistical summaries that describe central tendency and spread, along with correlation analysis to measure the strength of linear relationships. The summary emphasizes the importance of MLE in parameter estimation and the relationships between sample statistics.

Uploaded by

tresa
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
0% found this document useful (0 votes)
3 views15 pages

Statistical Estimation and MLE Analysis

The document discusses statistical estimation, focusing on Maximum Likelihood Estimation (MLE) and its application to various distributions such as Bernoulli, Poisson, Exponential, and Normal. It also covers statistical summaries that describe central tendency and spread, along with correlation analysis to measure the strength of linear relationships. The summary emphasizes the importance of MLE in parameter estimation and the relationships between sample statistics.

Uploaded by

tresa
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

Statistical Estimation, MLE, Summaries and

Correlation Analysis

Ajmal Mohammed

Department of ECE

September 23, 2025

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 1 / 15


Outline

1 Statistical Estimation

2 Maximum Likelihood Estimation (MLE)

3 Statistical Summaries

4 Correlation Analysis

5 Summary

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 2 / 15


Estimator

An estimator is a rule or formula used to estimate an unknown


population parameter.
Example:
n
1X
µ̂ = Xi estimates µ.
n
i=1

Properties: unbiasedness, consistency, efficiency.

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 3 / 15


MLE Concept

Given sample x1 , . . . , xn from distribution f (x; θ)


n
Y n
X
L(θ) = f (xi ; θ), ln L(θ) = ln f (xi ; θ)
i=1 i=1

MLE θ̂ maximizes ln L(θ).

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 4 / 15


Problem 1: Bernoulli Distribution

Problem: A coin is tossed n times with probability of head p. Derive the


MLE of p.
Solution:
P P X X
xi
L(p) = p (1 − p)n− xi
, ln L(p) = xi ln p + (n − xi ) ln(1 − p).

Differentiating:
P P
d xi n − xi

ln L(p) = = 0.
dp p 1−p
P
xi
⇒ p̂MLE = = X̄ .
n
Thus, the MLE of p is the sample mean.

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 5 / 15


Problem 2: Poisson Distribution

Problem: Let X1 , . . . , Xn ∼ Poisson(λ). Find the MLE of λ.


Solution:
n
Y e −λ λxi P Y 1
L(λ) = = e −nλ λ xi
xi ! xi !
i=1
X 
ln L(λ) = −nλ + xi ln λ + constant.
P P
d xi xi
ln L = −n + = 0 ⇒ λ̂ = = X̄ .
dλ λ n
So, MLE of λ is the sample mean.

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 6 / 15


Problem 3: Exponential Distribution
Problem: Let X1 , . . . , Xn ∼ Exponential(θ) with pdf

1
f (x; θ) = e −x/θ , x > 0.
θ
Find the MLE of θ.
Solution:  n P
1
L(θ) = e − xi /θ
θ
1X
ln L(θ) = −n ln θ − xi .
θ
Differentiating: P
d n xi
ln L = − + 2 = 0.
dθ θ θ
1X
⇒ θ̂ = xi = X̄ .
n
Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 7 / 15
Problem 4: Normal Distribution (Mean unknown, Variance
known)

Problem: X1 , . . . , Xn ∼ N(µ, σ 2 ) with known σ 2 . Find the MLE of µ.


Solution:
(xi − µ)2
 
1
f (xi ; µ) = √ exp −
2πσ 2 2σ 2
n 1 X
ln L(µ) = − ln(2πσ 2 ) − 2 (xi − µ)2 .
2 2σ
Differentiate wrt µ:
d 1 X
ln L = 2 (xi − µ) = 0 ⇒ µ̂ = X̄ .
dµ σ

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 8 / 15


Problem 5: Normal Distribution (Variance unknown)

Problem: X1 , . . . , Xn ∼ N(µ, σ 2 ) with mean µ known. Find the MLE of


σ2.
Solution:
n 1 X
ln L(σ 2 ) = − ln(2πσ 2 ) − 2 (xi − µ)2 .
2 2σ
Differentiate:
d n 1 X
ln L = − + (xi − µ)2 = 0.
dσ 2 2σ 2 2σ 4
1X
⇒ σ̂ 2 = (xi − µ)2 .
n

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 9 / 15


Statistical Summaries

1P
Mean: x̄ = xi
n
1 P
Variance: s 2 = (xi − x̄)2
n−1

Standard deviation: s = s 2

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 10 / 15


Problem: Mean and Variance

Problem: Find mean and variance for marks: 10, 20, 30, 40, 50.

10 + 20 + 30 + 40 + 50
x̄ = = 30
5
400 + 100 + 0 + 100 + 400
s2 = = 250, s ≈ 15.81.
4

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 11 / 15


Correlation

Measures linear relationship between x and y .


Pearson’s correlation coefficient:
P
(x − x̄)(y − ȳ )
r = pP
(x − x̄)2 (y − ȳ )2
P

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 12 / 15


Correlation: Detailed calculations (step-by-step)
Given data:
x = {2, 4, 6, 8, 10, 12}, y = {40, 55, 65, 70, 85, 95}.
Sample means:
x̄ = 7.00, ȳ = 68.333333 (approx 68.33).

x y x − x̄ y − ȳ (x − x̄)(y − ȳ ) (x − x̄)2 (y − ȳ )2

2 40 −5.00 −28.33 141.67 25.00 802.78


4 55 −3.00 −13.33 40.00 9.00 177.78
6 65 −1.00 −3.33 3.33 1.00 11.11
8 70 1.00 1.67 1.67 1.00 2.78
10 85 3.00 16.67 50.00 9.00 277.78
12 95 5.00 26.67 133.33 25.00 711.11

Sums: 370.00 70.00 1983.33

Now compute Pearson’s correlation coefficient r :


P
(x − x̄)(y − ȳ )
r = pP P .
(x − x̄)2 (y − ȳ )2

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 13 / 15


Correlation

Substitute the sums:


370.00 370.00
r = √ = √ .
70.00 × 1983.3333 138833.33
Evaluate:
√ 370.00
138833.33 ≈ 372.86 ⇒ r≈ ≈ 0.9930.
372.86

Interpretation: r ≈ 0.993 indicates a very strong positive linear


correlation between study hours and marks.

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 14 / 15


Summary

MLE is a powerful method of parameter estimation.


Many distributions (Bernoulli, Poisson, Exponential, Normal) have
simple MLEs linked to sample mean/variance.
Statistical summaries describe central tendency and spread.
Correlation quantifies strength of linear relationship.

Ajmal Mohammed (Department of ECE) Statistical Estimation September 23, 2025 15 / 15

You might also like