Bayesian Inductive Inference and Maximum Entropy

摘要: The principles of Bayesian reasoning are reviewed and applied to problems inference from data sampled Poisson, Gaussian Cauchy distributions. Probability distributions (priors likelihoods) assigned in appropriate hypothesis spaces using the Maximum Entropy Principle, then manipulated via Bayes’ Theorem. testing requires careful consideration prior ranges any parameters involved, this leads a quantitive statement Occam’s Razor. As an example general principle we offer solution important problem regression analysis; determining optimal number use when fitting graphical with set basis functions.