slide

Page 14 ... Optimality of scalar field Conjugate. Priors for Exp. Fam. 4. The discovery of Antidata and virtual data. 5. Optimality of Priors with tails following.
1MB taille 2 téléchargements 365 vues
R or L spinning?

We think through metaphors:

I have a brain I obs. x I want to understand x I need to predict future x’

How?

dataTheory no inmaculate obs. no theoretical vacuum no fact w/o. fiction no data w/o. theory.

Why ?

is a logical proposition in a domain of discourse

data must have meaning by

meaning

I mean

Theory

Theory

= explanation = compressing code = Probability distribution

obs.

hidden

The

tatistical Manifold

data manifold

finite measures

Sufficient map

Canonical Example:

Example: Car insurance.

Any!!

iid

Wrong!

Wrong!

Ignorance = Independence & Uniformity

spread

concentrate

What’s New? 1. Objective proc. for transforming prior info into prior distributions. 2. A new understanding of Data, Prior, and Likelihood. 3. Optimality of scalar field Conjugate Priors for Exp. Fam. 4. The discovery of Antidata and virtual data. 5. Optimality of Priors with tails following power laws. 6. Evaporation of the Bayesian/Freq. divide. 7. A dent at the Mind/Body problem. 8. A justification of Perelman’s Action. (That proved Thurston’s Geometrization Conjecture.) 9. A Geometric Theory of Ignorance. 10.The solution to a 260 year old problem: Objective Quantification of Ignorance in Statistical Inference.

Max (Ignorance) s.t. Whatever is known

(I forgot to mention that Bayes Theorem follows from this as a special case and in 2 very different ways!)

The Universe is a human thing