SLIDE 1
Jo Joint Probability Distribution
- Probability of any combination of features to
happen
Conditional Probability
‘Banana-shaped probability distribution’ Probability density function (PDF)
Conditional independence, Nave Bayes and Bayesian Networks Jo - - PowerPoint PPT Presentation
Conditional independence, Nave Bayes and Bayesian Networks Jo - - PowerPoint PPT Presentation
Fundamentals of AI Introduction and the most basic concepts Conditional independence, Nave Bayes and Bayesian Networks Jo Joint Probability Distribution Banana -shaped probability distribution Probability of any combination of
SLIDE 2
SLIDE 3
The story of Andrew (Moore) and Manuela
False True Event M Event S False True
SLIDE 4
SLIDE 5
SLIDE 6
SLIDE 7
0.18 0.42 0.12 0.28
Most probable
SLIDE 8
False True Event M Event S False True Event L False True
SLIDE 9
SLIDE 10
SLIDE 11
SLIDE 12
Event R False True
SLIDE 13
SLIDE 14
SLIDE 15
SLIDE 16
SLIDE 17
SLIDE 18
SLIDE 19
SLIDE 20
SLIDE 21
SLIDE 22
SLIDE 23
SLIDE 24
SLIDE 25
SLIDE 26
SLIDE 27
SLIDE 28
SLIDE 29
SLIDE 30
SLIDE 31
Example from real-life
SLIDE 32
Example from real-life
SLIDE 33
Example from real-life
SLIDE 34
Now, what is naïve Bayesian assumption?
- In simple words, it assumes that all variables (or a
set of variables) are all conditionally independent : the Bayesian net is not connected
- Or, we have an unconnected Bayesian net
connected to a single node
x y z t
C
x,y,z,t are conditionally independent given C This construction can be used to predict C from x,y,z,t values: this is Naïve Bayes classifier
SLIDE 35
What you should take with you
- Conditional independence of evens given other events
- Bayesian networks: convenient graphical way to
represent known causalities and compute joint probability distribution
- Naïve Bayesian assumption is the simplest case: we