What can we expect from neural network models?

Summary form only given. In each area of the brain, numerous neurons constitute elaborate networks. These networks are linked through numerous connections, and compose large-scaled neural systems. The neural systems generate major brain functions such as movement, cognition, emotion, and memory-learning. The brain as such has extensively been studied anatomically, physiologically and chemically, and our knowledge ever grows to cover every details of the brain. Important principles such as activity-dependent synaptic plasticity, multilayered integration of neuronal networks, modular organization of brain tissues have been revealed. Yet, shortage of knowledge is obvious when one tries to reproduce brain functions with models. While the simple perceptron model, adaptive filter model, feedforward adaptive control system model have successfully reproduced functions of the cerebellum, efforts to model other parts of the brain have met greater difficulties. It is still difficult to answer fundamental questions such as what is meant by the intricate structure of the basal ganglia? How the hippocampal circuit serves for cognitive memory and learning? How language is encoded within the neocortical network? Structures of motor, cognitive and emotional systems in the brain have been dissected to some extent, but it is yet difficult to figure out where and how our volition emerges, how we sense beauty, truth and virtue intuitively, and where in our brain consciousness resides. Complete understanding of the brain could be achieved only when one successfully models neural networks and systems so as to reproduce entire aspects of brain functions.