• DocumentCode
    808105
  • Title

    Neuro chips with on-chip back-propagation and/or Hebbian learning

  • Author

    Shima, Takeshi ; Kimura, Tomohisa ; Kamatani, Yukio ; Itakura, Tetsuro ; Fujita, Yasuhiko ; Iida, Tetsuya

  • Author_Institution
    Toshiba Res. & Dev. Center, Kawasaki, Japan
  • Volume
    27
  • Issue
    12
  • fYear
    1992
  • fDate
    12/1/1992 12:00:00 AM
  • Firstpage
    1868
  • Lastpage
    1876
  • Abstract
    A layered neural net realized with two chips is described. One chip implements 24×24 synapses, a local weight control mechanism, and quantized ±1 LSB, both momentum and weight update schemes. The other contains 24 neurons, implementing not only backward propagation (BP) but Hebbian learning, with 200-pF drive capability. Some experimental chip characteristics verifying the implemented techniques are given
  • Keywords
    Hebbian learning; VLSI; analogue processing circuits; backpropagation; neural chips; VLSI; analogue chip; drive capability; layered neural net; local weight control mechanism; momentum update; neuro chip; on-chip back-propagation; synapses; weight update; Algorithm design and analysis; Analog circuits; Hebbian theory; Learning systems; Limiting; Neural networks; Neurons; Signal processing algorithms; Very large scale integration; Voltage;
  • fLanguage
    English
  • Journal_Title
    Solid-State Circuits, IEEE Journal of
  • Publisher
    ieee
  • ISSN
    0018-9200
  • Type

    jour

  • DOI
    10.1109/4.173117
  • Filename
    173117