For me, when the parameters are fixed (deterministic) but unknown, it is easy to interpret. However in a Bayesian approach, say p(x=1)=p_1 (unknown). The parameter p_1 is stochastic, which has a probability distribution p(p_1). Then is this p(p_1) deterministic??? And why??? Of course I accept the idea... I mean, somewhere you gotta make assumptions. But the idea of assuming something to be deterministic or stochastic (just because...) somehow bugs me... Anyone got the same feelings?
[link][5 comments]