Bayesian Epistemology
2021; Cambridge University Press; Linguagem: Inglês
10.1093/obo/9780195396577-0417
ISSN1469-817X
Autores Tópico(s)Bayesian Modeling and Causal Inference
ResumoEveryone agrees that it is good to know. We however believe many more things than we know, which has made the notion of belief a target of recent work in epistemology. Not only do we believe propositions, we also believe them to different degrees. That beliefs come in degrees is often taken as a psychological fact and as a normative principle of rationality. The most prominent normative approach to beliefs which come in degrees is Bayesian epistemology. Bayesian degrees of belief are postulated to be represented by numbers in the unit interval [0, 1] obeying the axioms of probability. The convention is that a greater number expresses a stronger belief. The second postulate of Bayesian epistemology governs the change of beliefs whenever new evidence becomes available via updating procedures, Bayesian updating for categorical evidence and the more general Jeffrey updating for uncertain evidence.
Referência(s)