Statistics - Best Point Estimation


Advertisements

Point estimation involves the use of sample data to calculate a single value (known as a statistic) which is to serve as a "best guess" or "best estimate" of an unknown (fixed or random) population parameter. More formally, it is the application of a point estimator to the data.

Formula

${MLE = \frac{S}{T}}$

${Laplace = \frac{S+1}{T+2}}$

${Jeffrey = \frac{S+0.5}{T+1}}$

${Wilson = \frac{S+ \frac{z^2}{2}}{T+z^2}}$

Where −

  • ${MLE}$ = Maximum Likelihood Estimation.

  • ${S}$ = Number of Success .

  • ${T}$ = Number of trials.

  • ${z}$ = Z-Critical Value.

Example

Problem Statement

If a coin is tossed 4 times out of nine trials in 99% confidence interval level, then what is the best point of success of that coin?

Solution

Success(S) = 4 Trials (T) = 9 Confidence Interval Level (P) = 99% = 0.99. In order to compute best point estimation, let compute all the values −

Step 1

$ {MLE = \frac{S}{T} \\[7pt] \, = \frac{4}{9} , \\[7pt] \, = 0.4444}$

Step 2

$ {Laplace = \frac{S+1}{T+2} \\[7pt] \, = \frac{4+1}{9+2} , \\[7pt] \, = \frac{5}{11}, \\[7pt] \, = 0.4545}$

Step 3

$ {Jeffrey = \frac{S+0.5}{T+1} \\[7pt] \, = \frac{4+0.5}{9+1} , \\[7pt] \, = \frac{4.5}{10}, \\[7pt] \, = 0.45}$

Step 4

Discover Z-Critical Value from Z table. Z-Critical Value (z) = for 99% level = 2.5758

Step 5

$ {Wilson = \frac{S+ \frac{z^2}{2}}{T+z^2} \\[7pt] \, = \frac{4+\frac{2.57582^2}{2}}{9+2.57582^2} , \\[7pt] \, = 0.468 }$

Result

Accordingly the Best Point Estimation is 0.468 as MLE ≤ 0.5

Calculator

Useful Video Courses


Video

Class 11th Statistics for Economics

40 Lectures 3.5 hours

Madhu Bhatia

Video

Statistics

40 Lectures 2 hours

Megha Aggarwal

Video

Applied Statistics in Python for Machine Learning Engineers

66 Lectures 1.5 hours

Mike West

Video

An Introduction to Wait Statistics in SQL Server

22 Lectures 1 hours

Mike West

Video

Geospatial Data Science: Statistics and Machine Learning

Most Popular

60 Lectures 12 hours

Michael Miller

Video

Basic Statistics & Regression for Machine Learning in Python

65 Lectures 5 hours

Abhilash Nelson

Advertisements