Show Reference: "The Parameterless Self-Organizing Map Algorithm"

The Parameterless Self-Organizing Map Algorithm IEEE Transactions on Neural Networks, Vol. 17, No. 2. (March 2006), pp. 305-316, doi:10.1109/tnn.2006.871720 by Erik Berglund, Joaquin Sitte
@article{berglund-and-sitte-2006,
    abstract = {The parameterless self-organizing map ({PLSOM}) is a new neural network algorithm based on the self-organizing map ({SOM}). It eliminates the need for a learning rate and annealing schemes for learning rate and neighborhood size. We discuss the relative performance of the {PLSOM} and the {SOM} and demonstrate some tasks in which the {SOM} fails but the {PLSOM} performs satisfactory. Finally we discuss some example applications of the {PLSOM} and present a proof of ordering under certain limited conditions.},
    author = {Berglund, Erik and Sitte, Joaquin},
    citeulike-article-id = {542834},
    citeulike-linkout-0 = {http://dx.doi.org/10.1109/tnn.2006.871720},
    citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs\_all.jsp?arnumber=1603618},
    doi = {10.1109/tnn.2006.871720},
    institution = {Div. of Complex \& Intelligent Syst., Univ. of Queensland, St. Lucia, Australia},
    issn = {1045-9227},
    journal = {IEEE Transactions on Neural Networks},
    keywords = {learning, parameters, som, unsupervised-learning},
    month = mar,
    number = {2},
    pages = {305--316},
    posted-at = {2014-12-19 15:02:24},
    priority = {2},
    publisher = {IEEE},
    title = {The Parameterless {Self-Organizing} Map Algorithm},
    url = {http://dx.doi.org/10.1109/tnn.2006.871720},
    volume = {17},
    year = {2006}
}

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

In SOM learning, shrinking of the neighborhood size and decreasing update strength usually follow predefined schedules i.e. they only depend on the update step.

In the PLSOM algorithm, update strength depends on the difference between a data point and the best-matching unit's weight vector, the quantization error. A large distance, indicating a bad representation of that data point in the SOM, leads to a stronger update than a small distance. The distance is scaled relative to the largest quantization error encountered so far.

PLSOM reduces the number of parameters of the SOM algorithm from four to two.