Moment convergence in conditional limit theorems

From MaRDI portal





The sum of random variables \(Y_i\) conditioned on a given value of the sum of discrete random variables \(X_i\) is considered. \(X_i\) and \(Y_j\) are dependent, but the pairs \((X_i, Y_i)\) form an i.i.d. sequence. Let \((X_{ni}, Y_{ni})\) be i.i.d. copies of \(X_n\), \(Y_n\) and consider the triangular array with the sums \(S_{nN}= \sum^N_1 X_{ni}\), \(T_{nN}= \sum^N_1 Y_{ni}\). Convergence of the distribution of the conditioned normalized sum to a normal distribution, and convergence of its moments are proved. The limit normal distribution is only considered. The conditions in the theorems are rather complicated or ``not so elegant (author), but they are easily verified. Some applications are discussed, e.g.: the occupancy, the hashing with linear probing, the random forests, the branching processes, etc.











This page was built for publication: Moment convergence in conditional limit theorems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2748437)