Noise-enhanced clustering and competitive learning algorithms *Neural Networks*, Vol. 37 (January 2013), pp. 132-140, doi:10.1016/j.neunet.2012.09.012 by Osonde Osoba, Bart Kosko

@article{osoba-and-kosko-2013, abstract = {Noise can provably speed up convergence in many centroid-based clustering algorithms. This includes the popular k-means clustering algorithm. The clustering noise benefit follows from the general noise benefit for the expectationâ€“maximization algorithm because many clustering algorithms are special cases of the expectationâ€“maximization algorithm. Simulations show that noise also speeds up convergence in stochastic unsupervised competitive learning, supervised competitive learning, and differential competitive learning.}, author = {Osoba, Osonde and Kosko, Bart}, doi = {10.1016/j.neunet.2012.09.012}, issn = {08936080}, journal = {Neural Networks}, keywords = {clustering, competitive-learning, noise}, month = jan, pages = {132--140}, posted-at = {2013-02-18 10:37:17}, priority = {2}, title = {Noise-enhanced clustering and competitive learning algorithms}, url = {http://dx.doi.org/10.1016/j.neunet.2012.09.012}, volume = {37}, year = {2013} }

See the CiteULike entry for more info, PDF links, BibTex etc.

Noise can improve convergence in clustering algorithms.⇒

k-means is a special case of the EM algorithm⇒

"Stochastic competitive learning behaves as a form of adaptive quantization", because the centroids being adapted distribute themselves in the data space such that they minimize the quantization error (according to the distance metric being used).⇒