Best of MYX-news: close

Let's go to the Competition HUNINNO: 2014.X.15-17. Innovative stress-management and Virtual-Robot Farmer in Budapest
Special prize for the NGSTRESS-project in the finale!!!

Let's go to the Finale: Innovact Award 2014 (April 01-02, Reims)
Our prototype = Virtual Robot Farmer
based on similarity analyses

Special prizes from INNOREG-KMRIÜ (2013) for innovative Students
(Innovation Agency for Central-Hungary)
Dead heat for two tasks: quality management of studies and
new moduls for production plannig softwares
based on similarity analyses

National Innovation Prize of ITBN 2012 for SeaLog
a software for suspicion generating
based on similarity analyses

MY-X FREE, the online analytical tool
supported by INNOCSEKK programm in Hungary (2006-2009)
based on similarity analyses

'Ember-kísérletek': close2

avagy a prevenció robotizálása

Hová tart vajon egy-egy idősor?
avagy céginformációs rendszeradatok nagyító alatt

Milyen csoportok bújnak meg 2D-s pixel-halmazokban?
avagy mi az ideális klaszterszám?

Melyik honlap az ideális?
avagy komplex értékelési feladat éles szeműeknek

avagy egy kérdőív sok mindent elárulhat...

Last modified: 2015.07.19 * 14:08 - MIAÚ-RSS

Definitions of benchmarks for modelling experts on the field of forecasting, or appendix to the Sokal-phenomenon

Leading article: 2015. June (MIAU No. 202.)
(Previous article: MIAU No. 201.)

Keywords: theory of aimlessness, Occam's razor, planning, compatition, combinatorics, AAA

Abstract: In case of forecasts different indicators can be published (c.f. numeric fitting, fitting of parallelism of directions, etc.). But, the unique indicators can not inform the Readers about the holistic description of model fitting detailed enough. For example: if the category of the adult population in Europe should be estimated with a threshold of 100 cm, then a fitting quote about 95% seems to be too low, because the real ratio of the category distribution is over 99% - what can be approximated through a monotone strategy with a better fitting as before. Therefore without rational benchmarks it is not worth launching the creation of models (c.f. legal definition of new variety/sort of plants). Benchmarks can be defined through objective declarations, subjective declarations and based on antidiscriminative solutions...
More (DOC) *** More (PDF)

Please, send Your comments per email!