The general theory of stochastic population processes (Q2394758): Difference between revisions

In this comprehensive paper the author proposes to study the foundations of a general theory of stochastic population processes where population means a collection of individuals each of which may be found in any one state \(x\) of a fixed set \(X\) of states. The state of the population is given by the states \(x_1, \ldots, x_n\), \(n= 0, 1, 2, \ldots,\) where each of the \(x_\nu\) ranges over \(X\). One might, e.g., have a population whose individuals are characterized by their age, size, weight, etc. or a population of stars characterized by their mass, position, velocity, etc. or a population of galaxy clusters as studied by \textit{J. Neyman} and \textit{E. L. Scott} [Zbl 0085.42906]. It is assumed that there is a probability distribution \(P\) over some \(\sigma\)-field \(B\) of subsets of the space \(\mathfrak X\) of all population states. The author develops first the theory of such population probability spaces taking the individual state space \(X\) as an abstract space. Next he considers an alternative approach where the state of the population is characterized by an integral-valued function \(N\) on a class of subsets \(A\) of \(X\), \(N(A)\) being the number of individuals in the population with states in \(A\). He uses generating functionals introduced by D. G. Kendall [Zbl 0038.08803] and by M. S. Bartlett and D. G. Kendall [Zbl 0042.13802] which play a role similar to that of generating functions in standard probability. The theory is extended to multivariate populations and to population processes where the population state is a function of some independent variables. Eventually he extends the theory to the case of populations whose total size can be infinite with positive probability and in the last section interesting examples are studied.