About Unitary Rating Score Constructing
Machine Learning
2007-05-23 v1
Abstract
It is offered to pool test points of different subjects and different aspects of the same subject together in order to get the unitary rating score, by the way of nonlinear transformation of indicator points in accordance with Zipf's distribution. It is proposed to use the well-studied distribution of Intellectuality Quotient IQ as the reference distribution for latent variable "progress in studies".
Keywords
Cite
@article{arxiv.cs/0401005,
title = {About Unitary Rating Score Constructing},
author = {Kromer Victor},
journal= {arXiv preprint arXiv:cs/0401005},
year = {2007}
}
Comments
10 pages, 3 figures