REFINING EXPERT BASED EVALUATION ON THE BASIS OF A LIMITED QUANTITY OF DATA
Ukrainian Journal of Information Technology · 2019
DOI: 10.23939/ujit2019.01.058 ↗Kurz gefasst
Fachartikel in der Zeitschrift Ukrainian Journal of Information Technology, 2019. Verfasst von Yu. I. Hrytsiuk und 1 weiteren Autor:innen. Die Arbeit nennt 36 Literaturangaben.
Abstract (Originalsprache)
A technique has been developed to refine expert based evaluation of the probability distribution parameter of a random variable based on a limited amount of statistical data. This made it possible to identify the most informative data transmission channel (the most qualified expert) and get its reliable assessment. It has been established that the analysis and processing of a limited amount of data is carried out using well-known techniques in probability theory and mathematical statistics, where significant theoretical and practical experience has been accumulated. A mathematical model that describes the state of an object, process, or phenomenon is presented as a point estimate of the probability distribution parameter of a random variable, the value of which is obtained on the basis of a small sample of data. The modern approaches to the statistical estimation of a random variable are analyzed, the most common of which is the Bayesian approach. It is established that the most significant moment of the Bayesian estimation of the unknown parameter of the probability distribution of a random variable is the appointment of a certain function of the a priori density of its distribution. This function should correspond to the available preliminary information on the shape of the a priori probability distribution of this quantity. The traditional approach to identifying the most informative channel for transmitting data on the state of an object, the course of a process or phenomenon, and cutting off others is less reliable. This is carried out using the so-called mechanism of reducers of degrees of freedom. Its main disadvantage is that in the cut-off data transmission channels, there may be some useful information that is not involved in the development of an agreed solution. Therefore, it is necessary to introduce mechanisms of discriminators of degrees of freedom. They allow all data transmission channels to participate in the decision-making process in terms of importance, which corresponds to the greatest degree of their information content in the current situation. An illustrative example of the application of the considered methods of averaging data is shown, which reflects the results of calculations by iterations using the implementation mechanisms of both reducers and discriminators of degrees of freedom. These mechanisms reflect the features of the implementation of iterative algorithms that are characteristic of both methods of mathematical statistics and methods of a synergetic system of averaging data.
Zitieren
APA
Hrytsiuk, Y. I., & Ferneza, O. R. (2019). REFINING EXPERT BASED EVALUATION ON THE BASIS OF A LIMITED QUANTITY OF DATA. Ukrainian Journal of Information Technology. https://doi.org/10.23939/ujit2019.01.058
Vancouver
Hrytsiuk YI, Ferneza OR. REFINING EXPERT BASED EVALUATION ON THE BASIS OF A LIMITED QUANTITY OF DATA. Ukrainian Journal of Information Technology. 2019. doi:10.23939/ujit2019.01.058
BibTeX
@article{Hrytsiuk2019,
author = {Hrytsiuk, Yu. I. and Ferneza, O. R.},
title = {REFINING EXPERT BASED EVALUATION ON THE BASIS OF A LIMITED QUANTITY OF DATA},
journal = {Ukrainian Journal of Information Technology},
year = {2019},
doi = {10.23939/ujit2019.01.058}
}
EUVIMED