Transcript Recommendation of a Strategy

Chapter 4 - Random Variables
Todd Barr
22 Jan 2010
Geog 3000
Overview
➲


Discuss the types of Random Variables
Discrete
Continuous
➲
Discuss the Probability Density Function
➲
What can you do with Random Variables
Random Variables
➲
Is usually represented by an upper case X
➲
is a variable whose potential values are all
the possible numeric outcomes of an
experiment
➲
Two types that will be discussed in this
presentation are Discrete and Continuous
Discrete Random Variables
➲
Discrete Random Variables are whole numbers
(0,1,3,19.....1,000,006)
➲
They can be obtained by counting
➲
Normally, they are a finite number of values, but
can be infinite if you are willing to count that high.
Discrete Probability Distribution
➲
Discrete Probability Distributions are a description
of probabilistic problem where the values that are
observed are contained within predefined values
➲
As with all discrete numbers the predefined values
must be countable
➲
They must be mutually exclusive
➲
They must also be exhaustive
Discrete Probability Distro
Example
➲
The classic fair coin example is the best
way to demonstrate Discrete Probability
Classic Coin Example
➲
Experiment: Toss 2 Coins and Count the
Numbers of Tails
Physical Outcome
Value (x)
Probabilities, p(x)
Heads, Heads
0
¼ .25
Heads, Tails
Tails, Heads
1
½ .5
Tails, Tails
2
¼ .25
Classic Coin Continued
➲
Histogram of our tosses
0.6
0.5
0.4
0.3
0.2
0.1
0
0
1
2
Classic Coin Toss Summary
➲
Its easy to see from the Histogram on the previous
slide the area that each of the
results occupy
➲
If we repeat this test 1000 times, there is a strong
probability that our results will
resemble the previous Histogram but with some standard
variance
➲
For more on the classic coin toss and Discrete
Random Variables please go see the Educator video
at http://www.youtube.com/watch?v=T6eoHAjdAfM
Its all Greek to Me
➲
Mean and Variables of Random Variables,
symbology
➲
μ (Mu) is the symbol for population mean
➲
σ is the symbol for standard deviation
➲
s or x-bar are the symbols for data
Mean of Probability Distribution
➲
The Mean of Probability Distribution is a weighted
average of all the possible values within an
experiment
➲
It assists in controlling for outliers and its
important to determining Expected Value
Expected Value
➲
Within the discrete experiment, an expected value
is the probability weighted sums of all the potential
values
➲
Is symbolized by E[X]
Variance
➲
Variance is the expected value from the
Mean.
➲
The Standard Deviation is the Square Root
of the Variance
Continuous Random Variables
➲
Continuous Random Variables are defined by
ranges on a number line, between 0 and 1
➲
This leads to an infinite range of probabilities
➲
Each value is equally likely to occur within this
range
Continuous Random Variables
➲
Since it would nearly impossible to predict
the precise value of a CRV, you must
include it within a range.
➲
Such as, you know you are not going to get
precisely 2” of rain, but you could put a range at
Pr(1.99≤x≤2.10)
➲
This will give you a range of probability on the bell
curve, or the Probability Density Function
Probability Density Function
➲
The Probability Density of a Continuous Random
Variable is the area under the curve between points a
and n in your formula
➲
In the above Bell Curve the Probability Density
formula would be Pr(.57≤x≤.70)
➲
For a more detailed explanation please see the Khan
Academy at www.KhanAcademy.org
Adding Random Variables
➲
By knowing the Mean and Variance of of a
Random Variable you can use this to help predict
outcomes of other Random Variables
➲
Once you create a new Random Variable, you can
use the other Random Variables within your
experiment to develop a more robust test
➲
As long as the Random Variables are independent
this process is simple
➲
If the Random Variables are Dependent, then this
process becomes more difficult
Useful Links
➲
PatrickJMT
http://www.youtube.com/user/patrickJMT
➲
The Khan Academy
http://www.KhanAcademy.org
➲ Wikipedia
http://www.wikipeda.org