06 Moments

772 views

Published on

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
772
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
13
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

06 Moments

  1. 1. Stat310 Moments Hadley Wickham Saturday, 30 January 2010
  2. 2. Engineer Your Career Monday, February 15 7:00 PM - 8:30 PM McMurtry Auditorium Find out what you can do with a degree in engineering from a panel of successful Rice engineering graduates who have gone into a variety of professions. (Plus get dessert!) http://engineering.rice.edu/EventsList.aspx? EventRecord=13137 Saturday, 30 January 2010
  3. 3. Homework Due today. From now on, if late, put in Xin Zhao’s mail box in the DH mailroom. Another one due next Thursday Buy a stapler Use official name Saturday, 30 January 2010
  4. 4. 1. Finish off proof 2. More about expectation 3. Variance and other moments 4. The moment generating function 5. The Poisson distribution 6. Feedback Saturday, 30 January 2010
  5. 5. Proof, continued Saturday, 30 January 2010
  6. 6. Expectation of a function Saturday, 30 January 2010
  7. 7. Expectation Expectation is a linear operator: Expectation of a sum = sum of expectations (additive) Expectation of a constant * a function = constant * expectation of function (homogenous) Expectation of a constant is a constant. T 2.6.2 p. 95 Saturday, 30 January 2010
  8. 8. Your turn Write (or recall) the mathematical description of these properties. Work in pairs for two minutes. (Extra credit this week is to prove these properties) Saturday, 30 January 2010
  9. 9. Moments The ith moment of a random variable is defined as E(X i) = μ' . The ith central i moment is defined as E[(X - E(X)) i] = μ i The mean is the ________ moment. The variance is the ________ moment. Saturday, 30 January 2010
  10. 10. Name Symbol Formula 1 mean μ μ'1 2 variance σ2 μ2 = μ'2 - μ 2 3 skewness α3 μ3 /σ3 4 kurtosis α4 μ4 /σ4 Saturday, 30 January 2010
  11. 11. 3 4 var =1 0.6 0.5 0.4 skew = 0 0.3 kurt = 3.4 0.2 0.1 0.0 5 6 0.6 0.5 0.4 0.3 0.2 0.1 0.0 2 4 6 8 2 4 6 8 Saturday, 30 January 2010
  12. 12. 0.4 1.2 1.6 mean = 4 0.6 0.5 0.4 skew = 0 0.3 kurt ≈ 2.5 0.2 0.1 0.0 2.6 2.8 3.6 0.6 0.5 0.4 0.3 0.2 0.1 0.0 2 4 6 8 2 4 6 8 2 4 6 8 Saturday, 30 January 2010
  13. 13. −1.83 −1.03 −1.02 −0.91 0.8 0.6 0.4 0.2 0.0 −0.61 −0.21 0.21 0.91 0.8 0.6 0.4 0.2 0.0 1.02 1.83 0.8 0.6 mean ≈ 4 0.4 var = 1.3 0.2 0.0 2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8 Saturday, 30 January 2010
  14. 14. 1.00 1.46 1.59 mean = 4 0.5 0.4 skew = 0 0.3 var ≈ 4 0.2 0.1 0.0 1.87 2.05 2.26 0.5 0.4 0.3 0.2 0.1 0.0 2 4 6 8 2 4 6 8 2 4 6 8 Saturday, 30 January 2010
  15. 15. mgf The moment generating function (mgf) is Mx(t) = E(eXt) (Provided it is finite in a neighbourhood of 0) Why is it called the mgf? (What happens if you differentiate it multiple times). Useful property: If MX(t) = MY(t) then X and Y have the same pmf. Saturday, 30 January 2010
  16. 16. Plus, once we’ve got it, it can make it much easier to find the mean and variance Saturday, 30 January 2010
  17. 17. Expectation of binomial (take 2) Figure out mgf. (Random mathematical fact: binomial theorem) Differentiate & set to zero. Then work out variance. Saturday, 30 January 2010
  18. 18. Your turn Compute mean and variance of the binomial. Remember the variance is the 2nd central moment, not the 2nd moment. Saturday, 30 January 2010
  19. 19. Poisson 3.2.2 p. 119 Saturday, 30 January 2010
  20. 20. Poisson distribution X = Number of times some event happens (1) If number of events occurring in non- overlapping times is independent, and (2) probability of exactly one event occurring in short interval of length h is ∝ λh, and (3) probability of two or more events in a sufficiently short internal is basically 0 Saturday, 30 January 2010
  21. 21. Poisson X ~ Poisson(λ) Sample space: positive integers λ ∈ [0, ∞) Saturday, 30 January 2010
  22. 22. Examples Number of calls to a switchboard Number of eruptions of a volcano Number of alpha particles emitted from a radioactive source Number of defects in a roll of paper Saturday, 30 January 2010
  23. 23. Example On average, a small amount of radioactive material emits ten alpha particles every ten seconds. If we assume it is a Poisson process, then: What is the probability that no particles are emitted in 10 seconds? Make sure to set up mathematically. Saturday, 30 January 2010
  24. 24. mgf, mean & variance Random mathematical fact. Compute mgf. Compute mean & 2 nd moment. Compute variance. Saturday, 30 January 2010
  25. 25. Next week Repeat for continuous variables. Make absolutely sure you have read 2.5 and 2.6. (hint hint) Saturday, 30 January 2010
  26. 26. Feedback Saturday, 30 January 2010

×