Partition Mapping based on Abstract Approximation Spaces

Often an important factor affecting the performance of granule/interval based computing is the ability of the algorithm to efficiently transform the settings of one system to those of another. This paper presents a novel method to address this issue of mapping partitions from one n-dimensional space to another. The work builds upon our existing knowledge about Posets (and their algebra). This mapping methodology has been implemented within a rough neural network and the relevant results presented