In Chapter 4 it will be convenient to work in a general setting which covers both discrete and continuous-type probability distributions. To do this, we assume that probability distributions under considerations are given by their Radon–Nikodym derivatives with respect to underlying reference measures usually denoted by *μ* or *ν*. The role of a reference measure can be played by a counting measure supported by a discrete set or by the Lebesgue measure on ; we need only that the reference measure is locally finite (i.e. it assigns finite values to compact sets). The Radon–Nikodym derivatives will be called probability ...

Start Free Trial

No credit card required