Professional Documents
Culture Documents
Lecture 02
Lecture 02
SYSTEMS
Lecture 2
Outline
◦ This lecture is an introduction to elementary
probability theory
◦ Probability Model
◦ Categorize the observations
◦ Assign probabilities to the categories
◦ Axioms of Probability
◦ Corollaries
◦ Joint Probability
◦ Conditional Probability
Definition of Probability
◦ Probability is the measure of the likeliness that
an event will occur
◦ There is a 20% chance of rain
◦ Rain is an event
◦ 20% is a “measure of the likeliness” that rain will
occur (probability)
◦ Probability model:
◦ Outcome (𝝃) : 𝑢1 , 𝑢2 … 𝑢6
◦ An outcome cannot be further decomposed into other outcomes
◦ Sample space (𝑆): 𝑆 𝑢1, 𝑢2, 𝑢3, 𝑢4, 𝑢5, 𝑢6
◦ The sample space is the set of all possible outcomes, called S
◦ Events (𝐴,𝐵): Set of outcomes that are of interest to us
◦ A = {𝜉: such that user id is an even number}
◦ 𝐵 {𝜉: such that user id is greater than 4}
Probability Model
◦ In terms of a Venn diagram our experiment will look like
u5
u6
u2 u4
u3
u1
S
◦ So far we have just seen how to analyze the observations
from an experiment
◦ We have not assigned any probabilities yet
Probability Model
◦ Assigning probabilities
◦ Probability is assigned to outcomes and
events
2.
◦ The sample space includes all possible
outcomes. Probabilities are normalized so that
the maximum value is unity.
A u5
u1
u6
u2 u4
B
S u3
3.
Corollaries of Axioms of Probability
4. Discarding the condition of exclusivity i.e.
◦ If then
+ -
u5
u6
u2 u4
u1 u3
S
Corollaries of Axioms of Probability
Corollaries of Axioms of Probability
5. Partition Property: For any event , and disjoint
partition set we can write
Corollaries of Axioms of Probability
Corollaries of Axioms of Probability
6. If then
Assigning Probabilities
Assigning Probabilities
◦ Classical approach:
◦ atomic outcomes: decomposed events into simplest
events
◦ assigning equal probabilities to all atomic outcomes
in an experiment
◦ Probability assignment is
1
Pr 𝑢1 Pr 𝑢2 Pr 𝑢6
6
Pr 𝑢2 ∪ 𝑢4 ∪ 𝑢6
Pr 𝑢2 Pr 𝑢4 Pr 𝑢6 𝑏𝑦 𝐴𝑥𝑖𝑜𝑚 3
1 1 1
𝑏𝑦 𝑝𝑟𝑜𝑏𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑎𝑠𝑠𝑖𝑔𝑛𝑚𝑒𝑛𝑡
6 6 6
1/2.
Assigning Probabilities -
Classical approach
◦ Drawbacks:
◦ Not suitable for unequal probability assignment
u6
u4
B u2
S u3
u1 u3
S
u6
u4
B u2
S u3
u6
u2 u4
u1 u3
S
𝑛 ,
Pr 𝐴, 𝐵 lim
→ 𝑛
u6
◦ Classical Approach u2 u4
u5 Event B has u5
already
occurred
u6 u6
u4 => u2, u4, u3, u1 u2 u4
u2
cannot occur
u3 u3
u1 u1
S S
Conditional Probability
u5
u6
u2 u4
u3
S u1
◦ Rearranging yields
◦ Rearranging yields
(4)
(5)
Conditional Probability
◦ Useful for calculating certain conditional
probabilities
Implying
Example
◦ Assume that two bits are transmitted from the Tx to the Rx
◦ Bit is rcvd correctly (c), Bit is rcvd in error (e).
◦ 4 possible Outcomes: 𝑆 𝑒𝑒, 𝑒𝑐, 𝑐𝑒, 𝑐𝑐
Pr 𝑒𝑒 0.01, Pr 𝑒𝑐 0.01, Pr 𝑐𝑒 0.01, Pr 𝑐𝑐 0.97
◦ Find the probability of events
◦ A = “second bit is in error”
◦ B = “first bit is in error”
◦ conditional probability that the second bit is in error given that the first bit
is in error
◦ Solution
◦ Let 𝐴 denote the event that the second bit is in error
◦ 𝐴 𝑒𝑒, 𝑐𝑒
◦ Let B denote the event that the first bit is in error
◦ 𝐵 𝑒𝑒, 𝑒𝑐
Solution
◦ A “second bit is in error”
◦𝐴 𝑒𝑒, 𝑐𝑒
◦ 𝐴 is the union of two disjoint events (outcomes) 𝑒𝑒 and
𝑐𝑒
Pr 𝐴 Pr 𝑒𝑒 Pr 𝑐𝑒 0.02