Bottleneck networks were employed in order to estimate a low dimensional surface on which the high dimensional data lie. However, the fitting of closed surfaces like a sphere is hard for them. This is essentially due to the fact that general manifolds cannot be expressed by a single coordinate system. To overcome this difficulty, we multiply the bottleneck networks and construct a mixture of experts network, which can treat a broad class of surfaces. A procedure to restrain the premature convergence of the mixture of experts network is also provided.
The discovery of nitric oxide (NO) as an intercellular messenger brought about new concepts regarding the induction of synaptic plasticity in the neural circuits. Its importance has been recognized since the first enthusiasm, though certain inconsistent results have been reported recently on the effects of NO. In this review, latest studies are surveyed focusing on the functions of NO in vertebrate visual systems. There, this simple gas plays a role of adaptation signal in the retina, and it also acts as an important messenger involved in the self-organization of the optic nerve projection during the development.