- #36
julcab12
- 331
- 28
Torbjorn_L said:I'm no statistician (though I can use some basics). But I suspect that the above is not quite correct. Bayesian inferences are only promoted to probabilities that all agree on if the bets can be tested (HMM modeled).
Science use frequentist and bayesian probability methods of course, but mainly they use testing (as I described re the measurement theory that underlies it all), so likelihoods:
[ http://www-library.desy.de/preparch/books/vstatmp_engl.pdf ]
... In cosmology, Experimental means are quite limited. We cannot directly create, access,manipulate and infer a scenario with galactic bodies to extrapolate a better testing of the mechanics unlike local scientific methodology. Furthermore, regions in the universe are practically inaccessible by observation. So the only reliable approached is treat the dynamics--theory as an idealized and abstract mathematical model and materialize it by fitting the data using method of analysis and statistical techniques or approximation by prior assumption(like the cosmological principle) in the hope that it will be consistent all throughout. Or whether it can be tested through observations.
My point is, due to the limits of testing. We need an alternative reliable method in the Bayesian and statistical regime/spirit and clues in High energy physics to create an approximate judgement.