1. Set a. b. 2. 3. 4. Definitions a. Random Experiment: An experiment that can result in different outcomes, even though it is performed under the same conditions and in the same manner. b. Sample Space: This is the set of all possible outcomes. i. Discrete: Consists of finite or countable infinite set of outcomes ii. Continuous: Contains an interval (either finite or infinite) of real numbers c. Event: It is a subset of the sample space of a random experiment. i. Mutually Exclusive: ๐ธ1 ∩ ๐ธ2 = ∅ Probability a. ๐(๐ด′) = 1 − ๐(๐ด) b. ๐(๐ด ∪ ๐ต) = ๐(๐ด) + ๐(๐ต) − ๐(๐ด ∩ ๐ต) c. ๐(๐ด ∪ ๐ต ∪ ๐ถ) = ๐(๐ด) + ๐(๐ต) + ๐(๐ถ) − ๐(๐ด ∩ ๐ต) − ๐(๐ด ∩ ๐ถ) − ๐(๐ต ∩ ๐ถ) + ๐(๐ด ∩ ๐ต ∩ ๐ถ) d. ๐(๐ต) = ๐(๐ต ∩ ๐ด) + ๐(๐ต ∩ ๐ด′ ) Conditional Probability: ๐(๐ด|๐ต) → ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐ด ๐๐๐ฃ๐๐ ๐ต. a. 5. 6. ∪ → ๐๐๐๐๐ข๐: ๐ผ๐ก ๐๐ ๐๐๐๐๐๐๐ ๐ก๐ ๐กโ๐ "๐๐ " ∩ → ๐ผ๐๐ก๐๐๐ ๐๐๐ก๐๐๐: ๐ผ๐ก ๐๐ ๐๐๐๐๐๐๐ ๐ก๐ "๐ด๐๐ท" ๐(๐ต ∩ ๐ด) = ๐(๐ด|๐ต)๐(๐ต) โฏ ∴ โฏ ๐(๐ด|๐ต) = ๐(๐ต∩๐ด) ๐(๐ต) b. ๐(๐ต) = ๐(๐ต|๐ด)๐(๐ด) + ๐(๐ต|๐ด′)๐(๐ด′) → ๐๐จ๐ญ๐๐ฅ ๐๐ซ๐จ๐๐๐๐ข๐ฅ๐ข๐ญ๐ฒ ๐ซ๐ฎ๐ฅ๐: ๐ & ๐′ ๐๐ซ๐ ๐ฆ๐ฎ๐ญ๐ฎ๐๐ฅ๐ฅ๐ฒ ๐๐ฑ๐๐ฅ๐ฎ๐ฌ๐ข๐ฏ๐ Independence: Two events (A & B) are independent if one of the conditions listed below is satisfied; a. ๐(๐ด|๐ต) = ๐(๐ด) b. ๐(๐ต|๐ด) = ๐(๐ต) c. ๐(๐ด ∩ ๐ต) = ๐(๐ด)๐(๐ต) CIRCUIT PROBLEM:๐น๐๐ ๐๐๐๐๐ข๐๐ก๐ ๐๐ ๐ ๐๐๐๐๐ , ๐๐ข๐๐ก๐๐๐๐ฆ ๐กโ๐ ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐ ๐ข๐๐๐๐ ๐ ๐ก๐๐๐๐กโ๐๐. ๐(๐ท) = ๐(๐ด)๐(๐ต) = 0.992 = 0.98 A B ๐โ๐๐ ๐๐๐๐๐ก๐๐ ๐ก๐ ๐๐๐๐๐๐๐ก๐ ๐๐๐๐๐๐๐๐๐๐ “๐๐๐” 0.99 0.99 C 0.8 Fig 2:- Shows the probability of Success ๐น๐๐ ๐๐๐๐๐ข๐๐ก๐ ๐๐ ๐๐๐๐๐๐๐๐, ๐๐ข๐๐ก๐๐๐๐ฆ ๐กโ๐ ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐๐๐๐๐ข๐๐ ๐ก๐๐๐๐กโ๐๐. ๐กโ๐๐ ๐๐๐๐ ๐ ๐ข๐๐๐๐ ๐ ๐๐ฆ ๐ ๐ข๐๐ก๐๐๐๐ก๐๐๐ ๐กโ๐ ๐๐๐ ๐ค๐๐ ๐๐๐๐ 1. ๐(๐ธ) = 1 − (1 − ๐(๐ถ))(1 − ๐(๐ท)) = 1 − (0.2)(0.02) = 0.996 ๐โ๐๐ ๐๐๐๐๐ก๐๐ ๐ก๐ ๐๐๐๐๐๐๐ก๐ ๐๐๐๐๐๐๐๐๐๐๐๐ “๐๐” ๐(๐ธ) = ๐(๐ถ) + ๐(๐ท) − ๐(๐ถ)๐(๐ท) 7. a. b. c. d. 8. ๐(๐ด ∩ ๐ต) = ๐(๐ต ∩ ๐ด) = ๐(๐ต|๐ด)๐(๐ด) = ๐(๐ด|๐ต)๐(๐ต) ๐(๐ต|๐ด)๐(๐ด) = ๐(๐ด|๐ต)๐(๐ต) ๐(๐ด|๐ต )๐(๐ต) ๐(๐ต|๐ด) = ๐๐ผ๐๐ถ๐ธ ๐(๐ด) = ๐(๐ด|๐ต)๐(๐ต) + ๐(๐ด|๐ต′ )๐(๐ต′ ) ๐(๐ด) ๐(๐ด|๐ต )๐(๐ต) ๐(๐ต|๐ด) = ๐ & ๐ ′ ๐๐ซ๐ ๐ฆ๐ฎ๐ญ๐ฎ๐๐ฅ๐ฅ๐ฒ ๐๐ฑ๐๐ฅ๐ฎ๐ฌ๐ข๐ฏ๐ ′ ๐(๐ด|๐ต )๐(๐ต)+๐(๐ด|๐ต )๐(๐ต′ ) Mean (Expected Value): ๐ = ๐ธ(๐) = ∑ ๐ฅ๐(๐ฅ) ๐ฅ 9. Variance: ๐ 2 = ๐(๐) = ๐ธ(๐ − ๐)2 = ∑(๐ฅ − ๐)2 ๐(๐ฅ) = ∑ ๐ฅ 2 ๐(๐ฅ) − ๐ 2 ๐ฅ ๐ฅ 10. Standard Deviation: ๐ = √๐ 2 11. Discrete Uniform Distribution (DUD): This is when every member of the sample space has the same probability. ๐ผ๐ ๐ฅ1 , ๐ฅ2 , … , ๐ฅ๐ ๐๐จ๐ญ๐: if the values are not whole numbers, multiply by ๐๐๐ to make it whole. 1 ๐(๐ฅ๐ ) = ๐ a. For DUD, the ๐ฅ๐ + ๐ฅ1 ๐๐๐๐(๐) = ∗ 10−๐ 2 b. For DUD, the (๐ฅ๐ − ๐ฅ1 + 1)2 − 1 ๐ฝ๐๐๐๐๐๐๐(๐๐ ) = ∗ 10−2๐ 12 12. Bernoulli Trials:- A random experiment consists of Bernoulli trials if; The trials are independent. Each trial results in only two possible outcomes (Success and Failure). The probability of success in each trial remains constant a. Binomial Random Variable:- ๐ ๐๐๐๐๐ ๐ก๐ ๐ฅ − ๐๐ข๐๐๐๐ ๐๐ ๐ ๐ข๐๐๐๐ ๐ ๐๐ ๐๐ ๐ − ๐๐ข๐๐๐๐ ๐๐ ๐ก๐๐๐๐๐ A particularly long traffic light on your morning commute is ๐(๐ฅ) = ๐(๐ = ๐ฅ) ๐ฅ → ๐๐ข๐๐๐๐ ๐๐ ๐ ๐ข๐๐๐๐ ๐ ๐ ๐ฅ ๐−๐ฅ green 20% of the time that you approach it. Assume that each ๐ → ๐๐ข๐๐๐๐ ๐๐ ๐ก๐๐๐๐๐ = ( ) ๐ (1 − ๐) ๐ฅ morning represents an independent trial. ๐ → ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐ ๐ข๐๐๐๐ ๐ a) Over five mornings, what is the probability that the ๐๐๐๐(๐) = ๐ธ(๐) = ๐๐ light is green on exactly one day? ๐๐๐ก๐: ๐๐ข๐ ๐๐๐ ๐๐ ๐ค๐๐กโ๐๐ ๐ ๐๐๐๐๐ ๐ฅ = 1 → ๐๐๐ ๐๐๐ฆ ๐ = 5 → ๐๐๐ฃ๐ ๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐(๐ 2 ) ๐ท๐๐๐๐๐๐๐๐๐๐ ๐๐ ๐บ๐๐๐๐๐๐ ๐ = 0.2 → 20% = ๐(๐) = ๐๐(1 − ๐) Page 1 of 6 b. Geometric Random Variable:- ๐ ๐๐๐๐๐ ๐ก๐ ๐ฅ − ๐๐ข๐๐๐๐ ๐๐ ๐ก๐๐๐๐๐ ๐ข๐๐ก๐๐ ๐กโ๐ ๐๐๐๐๐ ๐ ๐ข๐๐๐๐ ๐ The probability of a successful optical alignment in the ๐(๐ฅ) = ๐(๐ = ๐ฅ) ๐ฅ → ๐๐ข๐๐๐๐ ๐๐ ๐ก๐๐๐๐๐ assembly of an optical data storage product is 0.8. Assume the = (1 − ๐) ๐ฅ−1 ∗ ๐ ๐ข๐๐ก๐๐ ๐๐๐๐ ๐ก ๐ ๐ข๐๐๐๐ ๐ 1 trials are independent. ๐ → ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐ ๐ข๐๐๐๐ ๐ ๐๐๐๐(๐) = ๐ธ(๐) = ๐ a) What is the probability that the first successful alignment requires exactly four trials? ๐๐๐ก๐: ๐๐ข๐ ๐๐๐ ๐๐ ๐ค๐๐กโ๐๐ ๐ ๐๐๐๐๐ ๐ฅ = 4 → ๐๐๐ข๐ ๐ก๐๐๐๐๐ ๐๐๐๐๐๐๐๐(๐ 2 ) = ๐(๐) 1−๐ ๐ = 0.8 = 2 ๐ c. Negative Binomial:- ๐ ๐๐๐๐๐ ๐ก๐ ๐ฅ − ๐๐ข๐๐๐๐ ๐๐ ๐ก๐๐๐๐๐ ๐ข๐๐ก๐๐ ๐ − ๐๐ข๐๐๐๐ ๐๐ ๐๐๐๐๐ข๐๐ ๐(๐ฅ) = ๐(๐ = ๐ฅ) ๐ → ๐๐ข๐๐๐๐ ๐๐ ๐๐๐๐๐ข๐๐ ≥ 1 ๐ฅ−1 ๐ ๐ฅ → ๐๐ข๐๐๐๐ ๐๐ ๐ก๐๐๐๐๐ ≥ 1 ๐ฅ−๐ =( ) ๐ (1 − ๐) ๐−1 ๐ → ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐๐๐๐๐ข๐๐ ๐ ๐๐๐๐(๐) = ๐ธ(๐) = ๐ ๐๐๐ก๐: ๐๐ข๐ ๐๐๐ ๐๐ ๐ค๐๐กโ๐๐ ๐ ๐๐๐๐๐ ๐๐๐๐๐๐๐๐(๐ 2 ) = ๐(๐) (1−๐)๐ ๐ท๐๐๐๐๐๐๐๐๐๐ ๐๐ ๐๐๐๐๐๐๐ = 2 ๐ 13. Hypergeometric: - ๐ ๐๐๐๐๐ ๐ก๐ ๐ ๐ ๐๐๐๐๐ ๐๐ ๐ − ๐๐๐ ๐ก๐๐๐๐๐ ๐๐ข๐ก ๐๐ ๐ ๐ ๐๐๐๐๐ ๐๐ ๐ − ๐๐๐ (๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐๐) ๐ค๐๐กโ ๐๐๐ก๐๐๐๐ ๐ก ๐๐ ๐ฅ − ๐๐ข๐๐๐๐ ๐๐ ๐๐๐๐๐๐ก๐ ๐๐๐๐ ๐. A state runs a lottery in which six numbers are randomly ๐(๐ฅ) = ๐(๐ = ๐ฅ) ๐ → ๐๐๐ก๐๐ # ๐๐ ๐๐๐. ๐−๐พ ๐พ selected from 40, without replacement. A player chooses six ๐ → ๐๐๐ก๐๐ # ๐ก๐๐๐๐๐ ๐๐ข๐ก ๐๐ ๐ ( )( ) ๐ − ๐ฅ ๐ฅ numbers before the state’s sample is selected. ๐พ → ๐๐๐ ๐ ๐๐๐๐ # ๐๐ ๐ ๐ข๐๐๐๐ ๐ ๐๐ ๐ = ๐ a) What is the probability that the six numbers chosen ๐ฅ → # ๐๐ ๐ ๐๐๐ ๐ค๐ ๐ค๐๐๐ก (๐พ ๐๐๐ ๐ก๐ฆ๐๐) ( ) ๐พ ๐ by a player match all six numbers in the state’s ๐ = → ๐๐๐๐๐๐๐ก๐๐๐ ๐๐ ๐ ๐ข๐. ๐๐ ๐ ๐ sample? ๐−๐ ๐๐๐๐(๐) = ๐ธ(๐) = ๐๐ ( ) → ๐๐ ๐กโ๐ ๐๐๐๐๐๐๐ก๐๐๐ ๐๐๐๐ก๐๐ 2) 6 ๐−1 ๐ = 40 ๐พ = 6 → ๐ ๐๐๐๐๐ ๐๐๐๐๐๐๐๐(๐ = ๐(๐) ๐ = = 0.15 40 ๐−๐ ๐ = 6 → ๐โ๐๐ ๐๐ ๐ฅ = 6 → ๐๐๐ก๐โ๐๐ = ๐๐(1 − ๐) ( ) ๐−1 14. Poisson Distribution:- This refers to any random experiment with the properties below; Counting the number of occurrence of an event over a period of time/space. The # of occurrence of an event in an interval is proportional to the length of the interval. Events cannot occur simultaneously. Occurrences of the events are independent for non-overlapping intervals. ๐(๐ฅ) = ๐(๐ = ๐ฅ) ๐ −๐๐ก (๐๐ก) ๐ฅ = ๐ฅ! ๐ → ๐ ๐๐ก๐ ๐๐ ๐๐๐๐ข๐๐๐๐๐๐ ๐ก → ๐โ๐ ๐๐๐๐ข๐๐ก ๐๐ ๐ก๐๐๐ ๐ข๐๐๐ก ๐ฅ → # ๐๐ ๐๐ฃ๐๐๐ก๐ ๐ = ๐๐ ๐๐๐๐(๐) = ๐ธ(๐) = ๐ ๐๐๐๐๐๐๐๐(๐ 2 ) = ๐(๐) = ๐ 15. Continuous Random Variable (C.R.V) ๐ฅ ๐น(๐ฅ) = ๐(๐ ≤ ๐ฅ) = ∫ ๐(๐ฅ)๐๐ฅ −∞ ๐ฅ ๐๐๐๐(๐) = ∫ ๐ฅ๐(๐ฅ)๐๐ฅ −∞ ๐๐๐๐๐๐๐๐(๐ 2) ๐ฅ = ∫ ๐ฅ 2 ๐(๐ฅ)๐๐ฅ − ๐ 2 −∞ 16. Continuous Uniform Distribution: ๐ผ๐ ๐ฅ1 , ๐ฅ2 , … , ๐ฅ๐ 1 ๐(๐ฅ๐ ) = ๐ฅ๐ − ๐ฅ1 ๐๐จ๐ญ๐: if the values are not whole numbers, multiply by ๐๐๐ to make it whole. a. For CUD, the b. ๐๐๐๐(๐) = ( For CUD, the ๐ฅ๐ +๐ฅ1 2 ๐๐๐๐๐๐๐๐(๐ 2 ) = ( ) ∗ 10−๐ (๐ฅ๐ −๐ฅ1 )2 12 ) ∗ 10−2๐ 17. Normal Distribution: (๐ฅ−๐)2 1 − ๐(๐ฅ) = ๐ 2๐2 √2๐๐ ๐ ๐๐ ๐๐๐ ๐๐๐๐ ๐๐๐๐๐ ๐๐ ๐๐ ๐๐๐ ๐๐๐๐๐๐๐๐. a. Note: you can only read from the table when ๐ = ๐ ๐๐๐ ๐ = ๐. standardized… ๐ผ๐ ๐ฟ ๐๐ ๐ ๐๐๐๐๐๐ ๐ฃ๐๐๐๐๐๐๐ (๐๐๐ก ๐ ๐ก๐๐๐๐๐๐)๐ค๐๐กโ ๐ฌ(๐) = ๐ ๐๐๐ ๐ฝ(๐) = ๐๐ , ๐กโ๐ ๐ ๐ ๐−๐ ๐= ๐๐ ๐ ๐ก๐๐๐๐๐๐ ๐๐๐๐๐๐ ๐ ๐ ๐ค๐๐กโ ๐ฌ(๐) = ๐, ๐ฝ(๐) = ๐. ๐ 18. Normal Approximation to Binomial Distribution: Use iff both ๐๐ > ๐ ๐๐๐ ๐๐(๐ − ๐) > ๐ ๐−๐ ๐−๐๐ a. ๐ผ๐ ๐ ๐๐ ๐ ๐๐๐๐๐๐๐๐ ๐๐๐ ๐ก. ๐ = = ๐ b. c. √๐๐(๐−๐) ๐๐ ๐ข๐ ๐ ๐กโ๐๐ , ๐ฆ๐๐ข ๐๐ข๐ ๐ก ๐ข๐ ๐ ๐กโ๐ ๐๐๐๐๐๐๐ก๐๐๐ ๐๐๐๐ก๐๐ ๐๐ 0.5. (๐. ๐. ๐๐๐ ≤ ๐, ๐ข๐ ๐ + ๐. ๐ ๐๐๐ ๐๐๐ ≥ ๐ ๐ข๐ ๐ − ๐. ๐ ๐+๐.๐−๐๐ ๐๐๐ (๐ ≤ ๐ฅ) ๐ข๐ ๐ ๐(๐ ≤ (๐ฅ + 0.5)) … ∴ ๐ท(๐ ≤ ๐) = √๐๐(๐−๐) Page 2 of 6 19. Normal Approximation to Poison RV: Use iff both ๐ > ๐ ๐±๐.๐−๐ ๐±๐.๐−๐ a. ๐ = = ๐ ๐๐ ๐๐๐๐ ๐๐ ๐๐๐ ± 0.5 ๐๐ ๐๐๐๐๐๐๐๐ ๐๐๐ ๐ก. ๐ √๐ 20. Exponential Distribution: Amount of wait time until first count is obtained use integral. ๐๐ ๐(๐) = ๐๐−๐๐ ๐๐๐ 0 ≤ ๐ฅ ≤ ∞ ๐ ๐ญ(๐) = ๐(๐ฅ ≤ ๐) = ∫0 ๐๐ −๐๐ฅ ๐๐ฅ = ๐ − ๐−๐๐ ๐๐๐๐๐ ๐๐ ๐๐๐๐๐๐ ๐๐๐๐๐. ∞ ๐๐๐๐(๐) = ๐ญ(๐) = ๐(๐ฅ ≥ ๐) = ∫ ๐๐ −๐๐ฅ ๐๐ฅ = ๐−๐๐ ๐ 1 = ๐ ๐๐๐ ๐ 2 = 2 ๐ ๐ ๐๐๐๐ ๐ข๐๐ก๐๐ ๐ ๐ข๐๐๐ก๐ = ๐(๐ฅ < ๐) = ∑๐−1 0 ๐ Lack of Memory Property: ๐โ๐๐๐; ′๐′ → ๐ก๐๐๐ ๐๐๐๐๐๐๐ฆ ๐๐๐ ๐ก ๐(๐ < (๐ + ๐)|๐ > ๐) ′๐ ′ → ๐๐ฅ๐ก๐๐ ๐ก๐๐๐ = ๐(๐ < ๐) −๐๐ ๐ → ๐๐๐ก๐ ๐๐ ๐๐๐๐ข๐๐๐๐๐๐ =๐−๐ ๐ ๐ ๐−๐๐ก (๐๐ก) ๐ฅ! ๐ฅ a. What is the probability that EQ is detected before 2 years if it doesn’t occur the first year? EQ occurs at a rate of 1.5 yearly. ๐ = 1, ๐ = 2 − 1 = 1, ๐ = 1.5 21. ๐ฎ๐๐๐๐ ๐ซ๐๐๐. : ๐ด๐๐๐ข๐๐ก ๐๐ ๐๐๐๐ ๐๐๐๐ ๐๐๐๐๐ ‘๐’ ๐๐๐๐๐๐ ๐๐ ๐๐๐ก๐๐๐๐๐ (๐ > 1). 3 1 1 a. ๐น๐๐ ๐๐๐๐๐ก๐๐๐๐ ; ๐ช(๐) = (๐ − ๐)๐ช(๐−๐) … → Γ ( ) = ( ) Γ ( ) 2 2 2 b. ๐๐๐ ๐ผ๐๐ก๐๐๐๐๐ ; ๐ช(๐) = (๐ − ๐)! … → Γ(5) = 4! = 24 ๐ ๐ ๐ต๐๐๐: ๐ช(๐) = ๐ ๐๐๐ ๐ช ( ) = √๐ … → ๐๐๐ ๐๐๐๐ ๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐ ๐๐ ๐๐๐๐๐๐ ๐๐ . ๐ ๐ ๐ข๐๐๐ก๐ a) What is the mean time until a packet is ๐๐ ๐ฅ ๐−1 ๐ −๐๐ฅ ๐ → ๐๐๐ก๐ ๐๐ ๐๐๐๐ข๐๐๐๐๐๐ ( ) ๐ด๐๐๐ ๐ญ๐๐๐๐๐๐๐; ๐(๐ฅ) = formed, that is, until five messages have ๐ก๐๐๐ Γ(๐) ๐ฅ → ๐๐๐๐๐๐ ๐๐๐๐๐ ๐๐๐๐๐ก๐๐๐ ๐๐ ๐. arrived at the node? ๐ช๐๐๐๐๐๐๐๐๐๐; b) What is the probability that a packet is ๐ → # ๐๐ ๐ข๐๐๐ก๐ ๐๐ ๐๐๐๐๐ก๐๐ ๐ก๐ ๐ ๐−1 −๐๐ฅ formed in less than 10 seconds? ๐ → # ๐๐ ๐ข๐๐๐ก๐ ๐ข๐ ๐๐ ๐๐ ๐ ๐๐๐๐๐๐๐ก๐๐๐ ๐ (๐๐ฅ)๐ ๐น(๐) = ๐(๐ > ๐) = ∑ 4-116. In a data communication system, several messages that arrive at a node are bundled into a packet before they are transmitted over the network. Assume the messages arrive at the node according to a Poisson process with ๐ = ๐๐ messages per minute. Five messages are used to form a packet. ๐! ๐=0 ๐ ๐ ๐ ๐ฝ๐๐๐๐๐๐๐ (๐๐ ) = 2 ๐ ๐ด๐๐๐(๐) = ๐ = 30 ๐๐๐ ๐ ๐๐๐๐ ๐๐๐๐ข๐ก๐๐ 10 ๐ ๐๐๐๐๐๐ ๐ฅ= 60 ๐๐๐๐ข๐ก๐๐ ๐ = (≥ 5) ๐=5 → ๐๐๐๐ ๐๐๐๐ก ๐ต 22. Weibull Dist. : ๐๐๐๐ ๐ข๐๐ก๐๐ ๐๐๐๐๐ข๐๐ ๐๐ ๐๐๐๐ฆ ๐๐๐๐๐๐๐๐๐ก ๐โ๐ฆ๐ ๐๐๐๐ ๐ ๐ฆ๐ ๐ก๐๐๐ . ๐ฟ๐๐๐ ๐๐๐ ๐ ๐๐๐/๐๐ ๐น. ๐ด๐๐๐ ๐ญ๐๐๐๐๐๐๐; (๐ฝ−1) ๐ฅ ๐ฝ ๐ฝ ๐ฅ ๐(๐ฅ) = [( ) ( ) ] [๐ −(๐ฟ) ] ๐ฟ ๐ฟ ๐ช๐๐๐๐๐๐๐๐๐๐; ๐น(๐ฅ) = ๐(๐ < ๐ฅ) = 1 − ๐ ๐(๐ > ๐ฅ) = ๐ฅ ๐ฝ −( ) ๐ฟ ๐ฅ ๐ฝ −( ) ๐ ๐ฟ ๐ด๐๐๐(๐) = ๐ฟΓ 1 (1+ ) ๐ฝ ๐ฝ๐๐๐๐๐๐๐ (๐๐ ) 2 = ๐ฟ 2Γ 2 (1+ ) ๐ฝ − ๐ฟ 2 [Γ 1 ] (1+ ) ๐ฝ ๐ฟ → ๐ด๐๐ค๐๐ฆ๐ ๐๐๐ฃ๐๐ ๐๐ ๐กโ๐ ๐๐ข๐๐ ๐ก๐๐๐ ๐ฝ → ๐ด๐๐ค๐๐ฆ๐ ๐๐๐ฃ๐๐ ๐๐ ๐กโ๐ ๐๐ข๐๐ ๐ก๐๐๐ ๐ฅ → ๐โ๐ ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐๐๐ ๐๐๐ ๐ช → ๐ป๐๐๐๐ ๐๐ ๐๐๐๐๐ ๐๐๐๐๐๐๐๐. 4-135. Suppose the lifetime of a component (in hours) is modeled with a Weibull distribution with ๐ท = ๐ and๐น = ๐๐๐๐. Determine the following: (a) ๐(๐ > 3000) (b) ๐(๐ > 6000 | ๐ > 3000) (c) value for x such that ๐(๐ > ๐ฅ) = 0.9 −( 3000 2 −( ) 4000 = 0.5698 ๐(๐ด ∩ ๐ต) ๐น๐๐ ๐๐๐๐ก ๐); ๐๐ ๐๐๐ ๐(๐ด|๐ต) = ๐(๐ต) ๐((๐ > 6000) ∩ (๐ > 3000)) = ๐(๐ > 6000) ๐((๐ > 6000) | (๐ > 3000)) = = 6000 2 ) 4000 3000 2 ) −( ๐ 4000 ๐ −( = 0.1054 0.5698 ๐(๐>6000) ๐(๐>3000) = 0.1850 ๐น๐๐ ๐๐๐๐ก ๐); ๐(๐ > ๐ฅ) = ๐ ๐ฝ −(๐ฅ) ๐ฟ ๐ฝ ๐ฅ = (− ln(๐(๐ > ๐ฅ))) (๐ฟ) ๐ฅ = (− ln(0.9))2 (4000) = (0.0111)(4000) = 44.4 โ๐๐ข๐๐ ๐น๐๐ ๐๐๐๐ก ๐); ๐ฅ = 3000 ๐(๐ > 3000) = 1 − (1 − ๐ ๐(๐ > 3000) = ๐ 3000 2 ) 4000 ) 23. Lognormal Dist. : ๐ฟ๐๐๐ ๐๐๐ θ ๐๐๐/๐๐ ๐. ๐ด๐๐๐ ๐ญ๐๐๐๐๐๐๐; (ln(๐ฅ)−๐)2 1 −( ) 2๐2 ๐(๐ฅ) = [ ] [๐ ] ๐ฅ๐√2๐ ๐ช๐๐๐๐๐๐๐๐๐๐; ln(๐ฅ) − ๐ ๐น(๐ฅ) = ๐(๐ ≤ ๐ฅ) = Φ [ ] ๐ ๐2 (๐+ ) 2 ๐ ๐ด๐๐๐(๐) = 2 2 ๐ฝ๐๐๐๐๐๐๐ (๐๐ ) = ๐ (2๐+๐ ) (๐ ๐ − 1) ๐ฅ → ๐โ๐ ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐๐๐ ๐๐๐ ๐ → ๐ด๐๐ค๐๐ฆ๐ ๐๐๐ฃ๐๐ ๐๐ ๐กโ๐ ๐๐ข๐๐ ๐ก๐๐๐ ๐ → ๐ด๐๐ค๐๐ฆ๐ ๐๐๐ฃ๐๐ ๐๐ ๐กโ๐ ๐๐ข๐๐ ๐ก๐๐๐ ๐ฝ → ๐ ๐๐๐๐๐ (๐๐๐๐๐ ๐ต๐๐๐๐๐ ๐๐๐๐๐) ๐); ๐ฅ = 13,300 ln(13300)−5 ๐(๐ < 13,300) = Φ [ ] 3 Φ[1.50] = 0.9332 ๐); 4-138. Suppose that X has a lognormal distribution with parameters ๐ = 5 and ๐2 = 9. Determine the following: (a) ๐(๐ < 13,300) (b) The value for ๐ฅ such that P(X ≤ x) = 0.95 (c) The mean and variance of ๐ Φ [ln ๐ฅ๐−๐] = 0.95 ( ) Φ[1.64] = 0.95 … ln(๐ฅ)−5 3 = 1.64 ๐ฅ = ๐ 3(1.65)+5 = 20,952 9 2 (5+ ) ๐๐๐ ๐๐ = ๐ (2(5)+9) (๐ 9 − 1) ๐ = 13,360 ๐๐๐ ๐๐ = 1.45 × 1012 ๐); ๐ = ๐ 24. Beta Dist. : ๐ฟ๐๐๐ ๐๐๐ ๐ ๐๐๐ ๐ท. ๐ด๐๐๐ ๐ญ๐๐๐๐๐๐๐; Γ(๐ผ+๐ฝ) ๐ผ−1 ๐(๐ฅ) = ๐ฅ (1 − ๐ฅ)๐ฝ−1 Γ(๐ผ) Γ(๐ฝ) ๐ช๐๐๐๐๐๐๐๐๐๐; ๐ฅ Γ(๐ผ+๐ฝ) ๐ผ−1 ๐(๐ < ๐ฅ) = ∫ ( ๐ฅ (1 − ๐ฅ)๐ฝ−1 ) ๐๐ฅ Γ(๐ผ) Γ(๐ฝ) 0 1 Γ(๐ผ+๐ฝ) ๐ผ−1 ๐(๐ > ๐ฅ) = ∫ ( ๐ฅ (1 − ๐ฅ)๐ฝ−1 ) ๐๐ฅ Γ(๐ผ) Γ(๐ฝ) ๐ฅ ๐ด๐๐๐(๐) = ๐ผ ๐ผ+๐ฝ ๐ผ๐ฝ ๐ฝ๐๐๐๐๐๐๐ = (๐ผ + ๐ฝ) 2 (๐ผ + ๐ฝ + 1) ๐ผ−1 ๐ด๐๐ ๐ = ๐ผ+๐ฝ−2 (๐๐ ) ๐ผ → ๐ด๐๐ค๐๐ฆ๐ ๐๐๐ฃ๐๐ ๐๐ ๐กโ๐ ๐๐ข๐๐ ๐ก๐๐๐ ๐ฝ → ๐ด๐๐ค๐๐ฆ๐ ๐๐๐ฃ๐๐ ๐๐ ๐กโ๐ ๐๐ข๐๐ ๐ก๐๐๐ ๐ช → ๐ป๐๐๐๐ ๐๐ ๐๐๐๐๐ ๐๐๐๐๐๐๐๐. ๐ฅ → ๐โ๐ ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ ๐๐๐ ๐๐๐ ๐ต๐๐๐: ๐ ≤ ๐ ≤ ๐ ๐๐๐๐๐๐. 4-150. Suppose X has a beta distribution with parameters ๐ถ = ๐ and ๐ท = ๐. ๐. Determine the following: (a) ๐(๐ < 0.25) (b) ๐(0.5 < ๐ ) (c) mean and variance ๐) ๐ฅ < 0.25 & ๐๐ข๐๐. 0.25 ๐(๐ < 0.25) = ∫0 ( Γ(5.2) Γ(1) Γ(4.2) ๐ฅ 0 (1 − ๐ฅ)3.2 ) ๐๐ฅ .25 ๐(๐ < 0.25) = (4.2)Γ(4.2) 3.2 ∫ (1 − ๐ฅ) ๐๐ฅ Γ(4.2) 0 ๐(๐ < 0.25) = 4.2 [− (1−๐ฅ) 4.2 0.25 4.2 ] 0 −4.2 4.2 (0.75 − 1) 4.2 = −(0.754.2 − 1) = 0.7013 ๐(๐ < 0.25) = ๐) ๐ฅ > 0.5 & ๐๐ข๐๐. 1 Γ(5.2) ๐(๐ > 0.5) = ∫ ( ๐ฅ 0 (1 − ๐ฅ)3.2 ) ๐๐ฅ Γ(1) Γ(4.2) 0.5 1 (4.2)Γ(4.2) 3.2 ๐(๐ > 0.5) = ∫(1 − ๐ฅ) ๐๐ฅ Γ(4.2) 0.5 ๐(๐ > 0.5) = 4.2 [− 4.2 (1 − ๐ฅ) 1 ] 4.2 0.5 −4.2 4.2 ๐(๐ ≤ 0.25) = (0.5 − 1) 4.2 = −(0 − 0.54.2 ) = 0.0544 1 4.2 ; ๐๐ = 5.2 6.2(5.2)2 ๐ = 0.1923 ; ๐๐ = 0.0251 0 ๐๐๐๐ = =0 5.2 − 2 ๐); ๐= Page 3 of 6 25. Joint Probability: ∞ ∞ ∫−∞ ∫−∞ ๐๐ฟ๐ (๐, ๐)๐ ๐ ๐ ๐ = ๐ ๐๐ฟ๐ (๐, ๐) ≥ ๐ ๐๐๐ ๐๐๐ ๐ฅ, ๐ฆ ๐บ๐๐๐๐๐๐ ๐๐๐ ๐: ๐ ∑∞ −∞ ๐๐ฟ๐ (๐, ๐) = ๐ ๐ผ๐๐ ๐๐๐๐๐ ๐๐๐๐ ๐๐๐๐๐ ๐๐๐๐๐๐ ๐๐๐๐′ ๐๐๐๐๐๐. (๐ฟ = ๐, ๐, ๐, … ๐ ๐๐๐ ๐ = ๐, ๐, ๐, … ๐) ๐ฅ๐ ๐ด๐๐๐ ๐ฌ(๐ฟ) = ∑(๐๐๐ × ๐ฅ๐ ) & ๐ธ(๐) = ∑[(๐. ๐)๐ × ๐ฅ๐ ] ๐ฅ0 ๐ฅ๐ ๐ฆ๐ ๐ฆ๐ ๐ฆ๐ ๐ฅ๐ ๐ฌ(๐ฟ๐) = ∑ [(๐ฅ๐ )(๐ฆ๐ )๐๐๐ ] ๐ฌ(๐) = ∑(๐๐๐ × ๐ฆ๐ ) & ๐ธ(๐) = ∑[(๐. ๐)๐ฆ × ๐ฆ๐ ] ๐ฆ0 ๐ฅ0 ๐ฅ๐ ๐ฆ0 ๐ฅ๐ ๐ฆ๐ ๐ฆ๐ 2 2 ๐ฝ๐๐๐๐๐๐๐ ๐ฝ(๐ฟ) = ∑ [(๐ฅ๐ − ๐ธ(๐)) × ๐๐๐ ] ๐ฝ(๐ฟ๐) = [๐ธ(๐๐) − ๐ธ(๐)๐ธ(๐)]2 ๐ฝ(๐) = ∑ [(๐ฆ๐ − ๐ธ(๐)) × ๐๐๐ ] ๐ฅ0 ๐ฆ0 ๐(๐, ๐) = โฌ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฅ ๐๐ฆ → ๐๐๐๐ ๐๐๐๐๐๐ ๐ ๐๐๐๐๐ ๐๐๐ ๐๐๐ ๐๐ ๐๐ ๐๐๐๐๐๐๐๐๐๐๐ ๐ ๐ถ๐๐๐ ๐ผ๐๐ ๐๐๐๐๐ ๐๐๐๐ ๐๐๐๐๐ ๐๐๐๐๐๐ (๐ < ๐ฟ < ๐ ๐๐๐ ๐ < ๐ < ๐ ) ∞ ∞ ∞ ๐ฌ(๐ฟ) = ∫ ∫ ๐ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฅ ๐๐ฆ ๐ด๐๐๐ −∞ −∞ ∞ ∞ −∞ −∞ ∞ ๐ฝ(๐ฟ) = ∫ ∫ ๐๐ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฅ ๐๐ฆ ๐ฝ๐๐๐๐๐๐๐ ∞ ∞ ๐ฌ(๐) = ∫ ∫ ๐ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฅ ๐๐ฆ −∞ −∞ ∞ ๐ฝ(๐) = ∫ ∫ ๐๐ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฅ ๐๐ฆ −∞ −∞ ∞ ๐ฌ(๐ฟ๐) = ∫ ∫ ๐๐ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฅ ๐๐ฆ ๐ฝ(๐ฟ๐) = [๐ธ(๐๐) − ๐ธ(๐)๐ธ(๐)]2 −∞ −∞ 1. Marginal Probability: ๐. ๐ ๐๐ข๐๐๐ก๐๐๐ ๐๐๐ ๐ ๐ฆ๐ ๐(๐ฅ) (๐) = ∑ ๐๐๐ (๐, ๐ฆ) → ๐ ๐๐๐๐๐๐ ๐ ๐ค๐๐กโ ๐ ๐ฆ0 ๐ฆ ๐๐ (๐ฅ) = ∫ ๐๐๐ (๐ฅ, ๐ฆ) ๐๐ฆ ๐ ๐ ๐๐๐๐๐ ๐ < ๐ < ๐ ∞ ๐(๐ < ๐ < ๐) = ∫ ∫ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฆ ๐๐ฅ ๐ −∞ ๐. ๐ ๐๐ข๐๐๐ก๐๐๐ ๐๐๐ ๐ ๐ฅ๐ ๐(๐ฆ) (๐) = ∑ ๐๐๐ (๐ฅ, ๐) → ๐ ๐๐๐๐๐๐ ๐ ๐ค๐๐กโ ๐ ๐ฅ ๐ฅ0 ๐๐ (๐ฆ) = ∫ ๐๐๐ (๐ฅ, ๐ฆ) ๐๐ฅ ๐ ๐ ๐๐๐๐๐ ๐ < ๐ < ๐ ๐(๐ฅ) (3) = 〈๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐ (๐ฅ = 3)〉 = 0.55 ๐(๐ฆ) (2) = 〈๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐ก๐ฆ ๐๐( ๐ฆ = 2)〉 = 0.25 ∞ ๐(๐ < ๐ < ๐) = ∫ ∫ ๐๐๐ (๐ฅ, ๐ฆ)๐๐ฅ ๐๐ฆ ๐ −∞ 2. Conditional Probability: ๐๐|๐ฅ (๐ฆ) = ๐๐๐ (๐ฅ, ๐ฆ) ๐๐ (๐ฅ) ∞ → ๐๐๐ฃ๐๐ ๐๐ข๐๐๐ก๐๐๐ ๐๐๐๐ ๐๐ฅ๐ → ๐ต๐ ๐๐๐๐๐๐๐๐๐๐๐ ๐๐๐๐ ๐๐ ๐๐๐๐ → ๐๐ข๐๐๐ก๐๐๐ ๐๐๐๐๐ฃ๐๐ ๐๐๐๐ฃ๐ ๐(๐ > ๐ต|๐ = ๐) = ∫ ๐๐|๐ฅ (๐ฆ) ๐๐ฆ → ๐๐๐ข๐ ๐๐ ๐ ๐๐๐ ๐๐๐ ๐ ๐กโ๐๐ ๐๐๐ก๐๐๐๐๐ก ๐กโ๐ ๐๐๐ ๐ก. ๐ต ๐ฌ(๐|๐) = ∫ ๐ ๐๐|๐ฅ (๐ฆ) ๐๐ฆ ๐ฌ(๐ฟ|๐) = ∫ ๐ ๐๐|๐ฅ (๐ฅ) ๐๐ฅ ๐ฆ ๐ท(๐ฟ = ๐|๐ = ๐) = ๐ฅ ๐(๐|๐ฅ) = [∫ ๐ฆ 2 ๐๐|๐ฅ (๐ฆ) ๐๐ฆ] − [๐ธ(๐|๐ฅ)]2 ๐ฆ ๐(๐|๐ฅ) = [∫ ๐ฆ 2 ๐๐|๐ฅ (๐ฆ) ๐๐ฆ] − [๐ธ(๐|๐ฅ)]2 ๐ฆ (๐ + ๐ + ๐)! ๐ ๐ท(๐ฟ = ๐, ๐ = ๐, ๐ = ๐) = ( ) (๐๐ )๐ (๐๐ ) (๐๐ )๐ ๐! ๐! ๐! Covariance (๐๐ฟ๐ ): ๐๐๐ = ๐ธ(๐๐) − ๐ธ(๐)๐ธ(๐) ๐ (๐ + ๐)! (๐๐ )๐ (๐๐ ) ๐! ๐! ๐ท(๐ = ๐) → ๐ผ๐๐ ๐๐๐๐๐๐๐๐ ๐ (๐ + ๐)! (๐๐ )๐ (๐๐ ) (๐๐ )๐ ๐! ๐! ๐ท(๐ฟ = ๐|๐ = ๐, ๐ = ๐) = ๐ท((๐ + ๐) = (๐ + ๐)) Correlation (๐๐ฟ๐ ): ๐๐๐ = ๐๐๐ ๐๐ ๐๐ Page 4 of 6 Page 5 of 6 Page 6 of 6