Analysis of bionic olfactory neural networks based on small-world networks view

To model mammalian olfactory neural systems, Kset models have been constructed by Prof. Walter J. Freeman. In K-set models, KIII model simulates the whole olfactory neural system and has novel characters different from conventional artificial neural networks, such as non-convergent ldquochaoticrdquo dynamics. Based on small-world networks view, the structural characteristics of KIII model are analyzed in this paper. Analytic results show some interesting results: (1) KIII model has large clustering coefficient; (2) there is the linear relationship between node number and characteristic path length in KIII model.