LAUSR

This article offers a method for building computer experiment designs. This method is based on the theory of marked point processes, otherwise known as object processes. Those designs rely on both the distribution of points in the experimental region and on some characteristics attached to those points, and are obtained using a Monte Carlo Markov chain method (MCMC). A detailed study about the convergence of the Markov chain has been carried out. Moreover, a comparison between our approach and the ones which is given by other existing computer designs has been drawn.