How Neural Computing Can Still Be Unconventional After All These Years

  • Michael A. Arbib
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4618)


Attempts to infer a technology from the computing style of the brain have often focused on general learning styles, such as Hebbian learning, supervised learning, and reinforcement learning. The present talk will place such studies in a broader context based on the diversity of structures in the mammalian brain – not only does the cerebral cortex have many regions with their own distinctive characteristics, but their architecture differs drastically from that of basal ganglia, cerebellum, hippocampus, etc. We will discuss all this within a comparative, evolutionary context. The talk will make the case for a brain-inspired computing architecture which complements the bottom-up design of diverse styles of adaptive subsystem with a top-level design which melds a variety of such subsystems to best match the capability of the integrated system to the demands of a specific range of physical or informational environments.

This talk will be a sequel to Arbib, M.A., 2003, Towards a neurally-inspired computer architecture, Natural computing, 2:1-46, but the exposition will be self-contained.


Basal Ganglion Reinforcement Learning Supervise Learning Broad Context Mammalian Brain 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Michael A. Arbib
    • 1
  1. 1.USC Brain Project, University of Southern California, Los Angeles, CA 90089-2520USA

Personalised recommendations