GeneRec

GeneRec is a generalization of the Recirculation algorithm, and approximates Almeida-Pineda recurrent backpropagation.[1][2] It is used as part of the Leabra algorithm for error-driven learning.[3]

The symmetric, midpoint version of GeneRec is equivalent to the contrastive Hebbian learning algorithm (CHL).[1]

See also

References

  1. 1 2 O'Reilly, R.C. Biologically Plausible Error-driven Learning using Local Activation Differences: The Generalized Recirculation Algorithm. Neural Computation, 8, 895-938. Abstract PDF
  2. ↑ GeneRec description in Computational explorations in cognitive neuroscience: understanding the mind by Randall C. O'Reilly,Yuko Munakata
  3. ↑ Leabra overview in Emergent
This article is issued from Wikipedia - version of the Wednesday, January 09, 2013. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.