Jump to content

Cluster-weighted modeling

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by Melcombe (talk | contribs) at 12:24, 7 July 2010 (clean citation). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

In statistics, cluster-weighted modeling (CWM) is an algorithm-based approach to density estimation in joint input-output space proposed by Neil Gershenfeld.[citation needed] The basic CWM algorithm gives a single output cluster for each input cluster. However, CWM can be extended to multiple clusters which are still associated with the same input cluster.[1] Each cluster in CWM is localized to a Gaussian input region, and this contains its own trainable local model.[2] It is recognized as a versatile inference algorithm which provides simplicity, generality, and flexibility; even when a feedforward layered network might be preferred, it is sometimes used as a "second opinion" on the nature of the training problem.[3]

The original form proposed by Gershenfeld describes two innovations:

  • Enabling CWM to work with continuous streams of data
  • Addressing the problem of local minima encountered by the CWM parameter adjustment process[3]

CWM can be used to classify media in printer applications, using at least two parameters to generate an output that has a joint dependency on the input parameters.[4]

References

  1. ^ Feldkamp, L.A. (2001). "Cluster-weighted modeling with multiclusters" (PDF). International Joint Conference on Neural Networks. 3 (1): 1710โ€“1714. {{cite journal}}: Unknown parameter |coauthors= ignored (|author= suggested) (help)
  2. ^ Boyden, Edward S. "Tree-based Cluster Weighted Modeling: Towards A Massively Parallel Real-Time Digital Stradivarius" (PDF). Cambridge, MA: MIT Media Lab. {{cite journal}}: Cite journal requires |journal= (help)
  3. ^ a b Prokhorov, A New Approach to Cluster-Weighted Modeling Danil V. "A New Approach to Cluster-Weighted Modeling". Dearborn, MI: Ford Research Laboratory. {{cite journal}}: Cite journal requires |journal= (help); Unknown parameter |coauthors= ignored (|author= suggested) (help)
  4. ^ Gao, Jun (2003-07-24). "CLUSTER-WEIGHTED MODELING FOR MEDIA CLASSIFICATION". Palo Alto, CA: World Intellectual Property Organization. {{cite journal}}: Cite journal requires |journal= (help); Unknown parameter |coauthors= ignored (|author= suggested) (help)