# Download Analytical Methods in Probability Theory: Proceedings of the by Harald Bergström (auth.), Daniel Dugué, Eugene Lukacs, Vijay PDF By Harald Bergström (auth.), Daniel Dugué, Eugene Lukacs, Vijay K. Rohatgi (eds.)

Read or Download Analytical Methods in Probability Theory: Proceedings of the Conference Held at Oberwolfach, Germany, June 9–14, 1980 PDF

Best probability books

Extra resources for Analytical Methods in Probability Theory: Proceedings of the Conference Held at Oberwolfach, Germany, June 9–14, 1980

Sample text

Hold spacings. ) sup 0

PROOF. a) We n e e d show subseauence (ii) x n ~ ~, gn(Xn) of the (iii) + ¢ '0 () xe integers when Xn ÷ x 0 and n ÷ ~ t h r o u g h for the cases: Xn + 0, a n X n ÷ ~' (iv) (i) x n ÷ x 0 E some (0,~), Xn ÷ x0 -< 0, a n X n --

This shows that J Since [~(t)] is ergodic, hence T is ergodic. To relate these results back to our original problem, define the real-valued function h on ~ by h(w(1),w (2)) = (w(1)(0),w~ 2)) = [ ~(0, w (i)) ,y0(w (2)) ]. Then it is straightforward to verify that h[T(w(1),w(2))] = [~(yl(w(2)),w(1)),yl(w(2))] or hT = [~(YI),YI] = (ZI,YI) , and in general hT n = [~(~n),Yn] = (Zn,Yn). Thus we have 22 THEOREM 2. d, positive random variables satisfying A2 and A3, then the [(Zn,Yn) ] process is erg0dic and we can estimate consistently all joint probabilities of the [~(t)] process b_~ means of observation of the [(Zn,Yn) ] process.