1 / 18

Introduction to Bayesian statistics

Introduction to Bayesian statistics. Three approaches to Probability Axiomatic Probability by definition and properties Relative Frequency Repeated trials Degree of belief (subjective) Personal measure of uncertainty Problems The chance that a meteor strikes earth is 1%

liang
Télécharger la présentation

Introduction to Bayesian statistics

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Introduction to Bayesian statistics • Three approaches to Probability • Axiomatic • Probability by definition and properties • Relative Frequency • Repeated trials • Degree of belief (subjective) • Personal measure of uncertainty • Problems • The chance that a meteor strikes earth is 1% • The probability of rain today is 30% • The chance of getting an A on the exam is 50%

  2. Problems of statistical inference • Ho: θ=1 versus Ha: θ>1 • Classical approach • P-value = P(Data | θ=1) • P-value is NOT P(Null hypothesis is true) • Confidence interval [a, b] : What does it mean? • But scientist wants to know: • P(θ=1 | Data) • P(Ho is true) = ? • Problem • θ “not random”

  3. Bayesian statistics • Fundamental change in philosophy • Θ assumed to be a random variable • Allows us to assign a probability distribution for θ based on prior information • 95% “confidence” interval [1.34 < θ < 2.97] means what we “want” it to mean: P(1.34 < θ< 2.97) = 95% • P-values mean what we want them to mean: P(Null hypothesis is false)

  4. Estimating P(Heads) for a biased coin • Parameter p • Data: 0, 0, 0, 1, 0, 1, 0, 0, 1, 0 • p = 3/10 = 0.3 • But what if we believe coin is biased in favor of low probabilities? • How to incorporate prior beliefs into model • We’ll see that p-hat = .22

  5. Bayes Theorem

  6. Example • Population has 10% liars • Lie Detector gets it “right” 90% of the time. • Let A = {Actual Liar}, • Let R = {Lie Detector reports you are Liar} • Lie Detector reports suspect is a liar. What is probability that suspect actually is a liar?

  7. More general form of Bayes Theorem

  8. Example • Three urns Urn A: 1 red, 1 blue Urn B: 2 reds, 1 blue Urn C: 2 reds, 3 blues • Roll a fair die. If it’s 1, pick Urn A. If 2 or 3, pick Urn B. If 4, 5, 6, pick Urn C. Then choose one ball. • A ball was chosen and it’s red. What’s the probability it came from Urn C?

  9. Bayes Theorem for Statistics • Let θ represent parameter(s) • Let X represent data • Left-hand side is a function of θ • Denominator on right-hand side does not depend on θ • Posterior distribution Likelihood x Prior distribution • Posterior dist’n = Constant x Likelihood x Prior dist’n • Equation can be understood at the level of densities • Goal: Explore the posterior distribution of θ

  10. A simple estimation example • Biased coin estimation: P(Heads) = p = ? • 0-1 i.i.d. Bernoulli(p) trials • Let be the number of heads in n trials • Likelihood is • For prior distribution use uninformative prior • Uniform distribution on (0,1): f(p) = 1 • So posterior distribution is proportional to f(X|p)f(p) = • f(p|X)

  11. Coin estimation (cont’d) • Posterior density of the form f(p)=Cpx(1-p)n-x • Beta distribution: Parameters x+1 and n-x+1 • http://mathworld.wolfram.com/BetaDistribution.html • Data: 0, 0, 1, 0, 0, 0, 0, 1, 0, 1 • n=10 and x=3 • Posterior dist’n is Beta(3+1,7+1) = Beta(4,8)

  12. Coin estimation (cont’d) • Posterior dist’n: Beta(4,8) • Mean: 0.33 • Mode: 0.30 • Median: 0.3238 • qbeta(.025,4,8), qbeta(.975,4,8) = [.11, .61] gives 95% credible interval for p • P(.11 < p < .61|X) = .95

  13. Prior distribution • Choice of beta distribution for prior

  14. Posterior Likelihood x Prior = [ px(1-p)n-x ] [ pa+1(1-p)b+1 ] = px+a+1(1-p)n-x+b+1 • Posterior distribution is Beta(x+a, n-x+b)

  15. Prior distributions • Posterior summaries: • Mean = (x+a)/(n+a+b) • Mode = (x+a-1)/(n+a+b-2) • Quantiles can be computed by integrating the beta density • For this example, prior and posterior distributions have same general form • Priors which have the same form as the posteriors are called conjugate priors

  16. Data example • Maternal condition placenta previa • Unusual condition of pregnancy where placenta is implanted very low in uterus preventing normal delivery • Is this related to the sex of the baby? • Proportion of female births in general population is 0.485 • Early study in Germany found that in 980 placenta previa births, 437 were female (0.4459) • Ho: p = 0.485 versus Ha: p < 0.485

  17. Placenta previa births • Assume uniform prior Beta(1,1) • Posterior is Beta(438,544) • Posterior summaries • Mean = 0.446, Standard Deviation = 0.016 • 95% confidence interval: [ qbeta(.025,438,544), qbeta(.975,438,544) ] = [ .415, .477 ]

  18. Sensitivity of Prior • Suppose we took a prior more concentrated about the null hypothesis value • E.g., Prior ~ Normal(.485,.01) • Posterior proportional to • Constant of integration is about 10-294 • Mean, summary statistics, confidence intervals, etc., require numerical methods • See S-script: http://www.people.carleton.edu/~rdobrow/courses/275w05/Scripts/Bayes.ssc

More Related