Download presentation
Presentation is loading. Please wait.
Published byReynard Hodges Modified over 9 years ago
3
A quick intro to Bayesian thinking
4
104
5
Frequentist Approach 10/14 Probability of 1 head next: = 0.714 0.714 X Probability of 2 heads next: = 0.51
6
Bayesian Approach Rev. T. Bayes ( 1707 - 1761 )
7
PosteriorLikelihoodPrior Normalising Constant
8
Frequentist: Bayesian: Likelihood(p) -> parameter p is a fixed constant. Posterior ∝ Likelihood(p) x Prior(p)
9
Beta Prior for p
10
Beta ∝ Binomial x Beta How would you bet ? Results : p = 48.5%
14
Example US election
15
Actual Results
19
Bayesian ●Prior knowledge ●Updatable with new information ●Data are fixed ●Parameters are described probabilistically ●Complex to compute Frequentist ●No prior information to model ●Data is a repeatable random sample ●Parameters are fixed ●Easy to compute
20
Questions? “Applying Bayesian functions ….” TNG - S4 Ep10
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.