@inproceedings{desieno-1988, address = {San Diego, CA}, abstract = {There are a number of neural networks that self-organize on the basis of what has come to be known as Kohonen learning. The author introduces a modification of Kohonen learning that provides rapid convergence and improved representation of the input data. In many areas of pattern recognition, statistical analysis, and control, it is essential to form a nonparametric model of a probability density function p(x). The purpose of the improvement to Kohonen learning presented is to form a better approximation of p (x). Simulation results are presented to illustrate the operation of this competitive learning algorithm}, author = {DeSieno, Duane}, booktitle = {Neural Networks, 1988., IEEE International Conference on}, doi = {10.1109/icnn.1988.23839}, institution = {HNC Inc., San Diego, CA, USA}, journal = {Proceedings of the International Conference on Neural Networks}, keywords = {learning, som}, month = jul, pages = {117--124}, posted-at = {2014-03-18 15:29:07}, priority = {2}, publisher = {IEEE}, title = {Adding a Conscience to Competitive Learning}, url = {http://dx.doi.org/10.1109/icnn.1988.23839}, year = {1988} }
See the CiteULike entry for more info, PDF links, BibTex etc.
One way of evening out distribution of SOM units in data space is using `conscience': a value which increases every time a neuron is BMU and decreases whenever it isn't. High conscience values then lead to a lower likelihood of being selected as BMU.⇒