Applying set theory to probability Dr. Ahmed Elmoasry Applying set theory to probability The mathematics we study is a branch of measure theory Probability is a number that describes a set Flip a coin. Did it land a head or tails? Give a lecture. How many students are seated in the second row? Walk to a bus stop. How long do you wait for the arrival of a bus? Applying set theory to probability Waiting time at a bus stop, you may consider: The time of the day The speed of each car The wait, the horsepower, and gear of each kind of bus used by the bus company. The work schedule of each bus driver. (some drivers drive faster) Example 1 An experiment consist of the following procedure, observation and model: Procedure: Flip a coin and let it lands on a table. Observation: observe which side (head or tail) faces you after the coin lands. Model: Heads and tails are equally likely. Note Experiment consist of both procedure and observations. Two experiments with the same procedure but with different observations are different experiments: Ex. 1.2: Flip a coin three times. Observe the sequence of heads and tails. Ex. 2.3: Flip a coin three times. Observe the number of heads. Outcome, sample space , event An Outcome of an experiment is any possible observation of the experiment. The Sample Space of an experiment is the finestgrain, mutually exclusive, collectively exhaustive set of all possible outcomes. An Event is a set of outcomes of an experiment. Ex. 1.1 S={h, t} Ex. 1.2: S={hhh, hht, hth, thh, htt, tht,tth,ttt} Ex. 1.3: S={0,1,2,3} Comparison between Set & probability Set Algebra Set Universal Set Element Probability Event Sample Space Outcome Example Suppose we roll a six sided die and observe the number of dots on the side facing upwards. We can label these outcomes i=1,2,…, 6 Sample Space is Example Subsets of S are E1={Roll 4 or higher}={ } E2={Roll is even}={ } E3= {the roll is square of an integer}= { } Probability Axioms A probability measure P[ ] is a function that maps events in the sample space to real number such that:Axiom 1: for any events A , P[A] 0. Axiom 2: P[S] =1. Axiom3 : A1,A2,A3,…. & Ai∩Aj= P[A1A2 …]=p[A1]+p[A2]+p[A3]+…. Theorem Theorem Theorem Theorem Equally Likely outcomes Theorem S={s1,s2,s3,…,sn}, each outcomes is equally likely then p[si] = 1/n 1 i n. Example: for a die S= {1,2,3,4,5,6} E1={4,5,6} p[E1]= E2={2,4,6} p[E1]= E3={1,4} p[E1]= Theorem The probability function (measure ) p[] satisfies:p[]= p[Ac]= A,B & A and B are disjoint , i.e. A∩B= p[AB]=p[A]+p[B]-p[A∩B] if AB , p[A] p[B] Theorem For any event A and event space {B1,B2,…,Bm}P[A]= p[A∩Bi] (AUB)c=Ac∩Bc ≠ Example A company has model of telephone usage. It classifies all calls as either long (L), if they last more than 2 minutes, or brief (B). It also observes whether calls carry voice (V) , data (D), or fax(F) The sample space S={} V F D L 0.3 0.15 0.12 B 0.2 0.15 0.08