Reconstructing settlement evolution from neolithic to Shang dynasty in Songshan mountain area of central China based on self-organizing feature map

The Self-Organizing Feature Map (SOFM) is one of the most popular neural network models, recently also adopted in archaeology to improve and enhance, on the basis of the availability of information and archaeological records, our understanding of the long-term human settlements and their evolution. In this paper, SOFM has been applied to classify prehistoric settlement size-grade in the Songshan Mountain Region in China, mainly focusing on the following four periods: Peiligang (9000-7000aBP), Yangshao (7000-5000aBP), Longshan (5000-4000aBP) and Xia-Shang (4000-3000aBP). Outputs from the SOFM analysis enabled us to capture the spatial relation between higher and lower grade settlements and to identify specific morphological patterns. This brought new light on the human settlements and their evolution in relations with the nature, environmental features, and cultural attitude in the Songshan Mountain Region where the Chinese civilization emerged and developed.