An application of multi-stochastic hyper-matrices to the generation of simulated monthly rainfall totals
by
Julia Piantadosi
University of South Australia
Coauthors: Phil Howlett

We show how an empirically derived multi-stochastic hyper-matrix can be used to describe the observed statistics for a finite sequence of randomly generated numbers. In particular we show that a hyper-matrix of dimension m can be used to describe the observed relative frequency of occurrence of all subsequences of length m. We consider an application of this methodology to the generation of simulated monthly rainfall totals at a particular location.

Date received: January 6, 2010