ML Anthology

The Bootstrap Widrow-Hoff Rule as a Cluster-Formation Algorithm

Geoffrey E. Hinton, Steven J. Nowlan
NeCo 1990 pp. 355-362
doi:10.1162/NECO.1990.2.3.355 /neco/1990/hinton1990neco-bootstrap/

Abstract

An algorithm that is widely used for adaptive equalization in current modems is the “bootstrap” or “decision-directed” version of the Widrow-Hoff rule. We show that this algorithm can be viewed as an unsupervised clustering algorithm in which the data points are transformed so that they form two clusters that are as tight as possible. The standard algorithm performs gradient ascent in a crude model of the log likelihood of generating the transformed data points from two gaussian distributions with fixed centers. Better convergence is achieved by using the exact gradient of the log likelihood.

NeCo Semantic Scholar

Cite

Text

Hinton and Nowlan. "The Bootstrap Widrow-Hoff Rule as a Cluster-Formation Algorithm." Neural Computation, 1990. doi:10.1162/NECO.1990.2.3.355

Markdown

[Hinton and Nowlan. "The Bootstrap Widrow-Hoff Rule as a Cluster-Formation Algorithm." Neural Computation, 1990.](https://mlanthology.org/neco/1990/hinton1990neco-bootstrap/) doi:10.1162/NECO.1990.2.3.355

BibTeX

@article{hinton1990neco-bootstrap,
  title     = {{The Bootstrap Widrow-Hoff Rule as a Cluster-Formation Algorithm}},
  author    = {Hinton, Geoffrey E. and Nowlan, Steven J.},
  journal   = {Neural Computation},
  year      = {1990},
  pages     = {355-362},
  doi       = {10.1162/NECO.1990.2.3.355},
  volume    = {2},
  url       = {https://mlanthology.org/neco/1990/hinton1990neco-bootstrap/}
}