Last modified: 2015.07.19 * 14:08 - MIAÚ-RSS
Robot-Eye, or approximation of the ideal number in the cluster analyse based on similarity analyses
Leading article: 2016. January (MIAU No. 209.)
(Previous article: MIAU No. 208.)
Keywords: robot-eye, hermeneutics, anti-discrimination models, optimizing, automation, artificial intelligence, similarity analysis
Abstract: The ideal number of clusters in case of a given data asset seems to have a marginal position in the literature published in Hungarian
based on the amount of Google results. Some of them deliver hardly any relevant information.
Other sources evaluate the problem as not to solve. If ideal sets of data
(like rectangular, triangular clouds or circle of coordinates) got analysed the k-mean solution produced anomalies.
These anomalies have massive impacts in case of arbitrary data assets.
With other words: the ideal number of clusters can only be interpreted in form of approximations.
Anomalies could also be detected through questionnaires, where human eyes and brains got used for interpretation tasks.
The objective of this article is, to report about the possibility and technical solution of a robot eye.
Multi-layered antidiscriminative similarity analyses(COCO-Y0) is capable, to ensure rational interpretation of pictures.
More (DOC) *** More (PDF)
Please, send Your comments per email!