A Silicon Primitive for Competitive Learning

Abstract

Competitive learning is a technique for training classification and clustering networks. We have designed and fabricated an 11- transistor primitive, that we term an automaximizing bump circuit, that implements competitive learning dynamics. The circuit per(cid:173) forms a similarity computation, affords nonvolatile storage, and implements simultaneous local adaptation and computation. We show that our primitive is suitable for implementing competitive learning in VLSI, and demonstrate its effectiveness in a standard clustering task.

Cite

Text

Hsu et al. "A Silicon Primitive for Competitive Learning." Neural Information Processing Systems, 2000.

Markdown

[Hsu et al. "A Silicon Primitive for Competitive Learning." Neural Information Processing Systems, 2000.](https://mlanthology.org/neurips/2000/hsu2000neurips-silicon/)

BibTeX

@inproceedings{hsu2000neurips-silicon,
  title     = {{A Silicon Primitive for Competitive Learning}},
  author    = {Hsu, David and Figueroa, Miguel and Diorio, Chris},
  booktitle = {Neural Information Processing Systems},
  year      = {2000},
  pages     = {713-719},
  url       = {https://mlanthology.org/neurips/2000/hsu2000neurips-silicon/}
}