Cellular Automata Based Neural Networks For Modelling Dual Complex Systems: Land-Use/Cover And Transport Networks

When we take the time to observe nature, we often are surprised by the complex forms and behaviours that may emerge, surprised by the self reproduction and the self organisation phenomena that we notice, and we are compelled to acknowledge how difficult it is to understand this complex nature in perpetual state of evolution. By analogy to this nature, this paper deals with possible spatial evolution by combining two spatial systems: land use and the transport systems. However, underlying this evolution, there are processes that cannot be easily captured through the use of simple methods/statistical tools. Nevertheless, thanks to the emergence in the past few decades of artificial intelligence based models such as cellular automata and neural network models, this paper proposes to simulate future spatial evolution in Luxembourg and the Greater region up to 2020. The simulation results will act as the knowledge base which will make it possible to understand the functioning of complex cross-border areas.