Comparative analysis of the regression methods and data group accounting method in modeling mineral processing
The relevance of the discussed issue is caused by the need to conduct a comparative analysis of regression methods and the data group accounting method in order to evaluate the effectiveness of their use in modeling mineral processing. The main aim of the study is to assess the effectiveness of appl...
Main Authors: | , |
---|---|
Format: | Article |
Language: | Russian |
Published: |
Tomsk Polytechnic University
2016-08-01
|
Series: | Известия Томского политехнического университета: Инжиниринг георесурсов |
Subjects: | |
Online Access: | http://izvestiya.tpu.ru/archive/article/view/1734 |
_version_ | 1797813067757649920 |
---|---|
author | Seyran Shamirovich Balasanyan Hermine Mikhaylovna Gevorgyan |
author_facet | Seyran Shamirovich Balasanyan Hermine Mikhaylovna Gevorgyan |
author_sort | Seyran Shamirovich Balasanyan |
collection | DOAJ |
description | The relevance of the discussed issue is caused by the need to conduct a comparative analysis of regression methods and the data group accounting method in order to evaluate the effectiveness of their use in modeling mineral processing. The main aim of the study is to assess the effectiveness of application of the regression methods and the data group accounting method in constructing statistical models of minerals processing based on the results of theoretical studies, computer simulation experiments and practical application of these methods. The methods used in the study: methods of mathematical statistics, simulation method, method of inductive modeling. The results. The authors have carried out thecomparative analysis of regression methods and the data group accounting method by theoretical investigations, computer simulations and practical applications at the construction of model describing the statistical dependence of profit on the output interval characteristics of the ore grinding technological system of Zangezur Copper and Molybdenum Combine. As a result of the logical analysis of the above mentioned methods the authors concluded that a relatively high predictive property of the models constructed by the data group accounting method is ensured both by the selection of the optimal model structure, and by description of the random error. Using the computer simulation experiments the authors investigated the influence of sample size, level of statistical noise on the predictive ability of the models built using both methods. It was ascertained that a relatively high predictive ability of models constructed by the data accounting group method occurs especially at moderate statistical noise and small samples, comparable to the number of input variables. The possibilities of these methods were studied as well in terms of identifying physical and systemic regularities of various objects with the specified postulated functions. The effectiveness of the practical application of the considered methods is evaluated by the results of construction of techno-economic model of ore grinding technological system. The application of a stepwise regression method allowed constructing the best possible model in terms of a compromise between adequacy and complexity. This indicates the feasibility of applying the regression methods at constructing the statistical models of minerals processing. |
first_indexed | 2024-03-13T07:47:53Z |
format | Article |
id | doaj.art-579f046805cc44af90999360d25b7744 |
institution | Directory Open Access Journal |
issn | 2500-1019 2413-1830 |
language | Russian |
last_indexed | 2024-03-13T07:47:53Z |
publishDate | 2016-08-01 |
publisher | Tomsk Polytechnic University |
record_format | Article |
series | Известия Томского политехнического университета: Инжиниринг георесурсов |
spelling | doaj.art-579f046805cc44af90999360d25b77442023-06-02T21:12:37ZrusTomsk Polytechnic UniversityИзвестия Томского политехнического университета: Инжиниринг георесурсов2500-10192413-18302016-08-013274Comparative analysis of the regression methods and data group accounting method in modeling mineral processingSeyran Shamirovich BalasanyanHermine Mikhaylovna GevorgyanThe relevance of the discussed issue is caused by the need to conduct a comparative analysis of regression methods and the data group accounting method in order to evaluate the effectiveness of their use in modeling mineral processing. The main aim of the study is to assess the effectiveness of application of the regression methods and the data group accounting method in constructing statistical models of minerals processing based on the results of theoretical studies, computer simulation experiments and practical application of these methods. The methods used in the study: methods of mathematical statistics, simulation method, method of inductive modeling. The results. The authors have carried out thecomparative analysis of regression methods and the data group accounting method by theoretical investigations, computer simulations and practical applications at the construction of model describing the statistical dependence of profit on the output interval characteristics of the ore grinding technological system of Zangezur Copper and Molybdenum Combine. As a result of the logical analysis of the above mentioned methods the authors concluded that a relatively high predictive property of the models constructed by the data group accounting method is ensured both by the selection of the optimal model structure, and by description of the random error. Using the computer simulation experiments the authors investigated the influence of sample size, level of statistical noise on the predictive ability of the models built using both methods. It was ascertained that a relatively high predictive ability of models constructed by the data accounting group method occurs especially at moderate statistical noise and small samples, comparable to the number of input variables. The possibilities of these methods were studied as well in terms of identifying physical and systemic regularities of various objects with the specified postulated functions. The effectiveness of the practical application of the considered methods is evaluated by the results of construction of techno-economic model of ore grinding technological system. The application of a stepwise regression method allowed constructing the best possible model in terms of a compromise between adequacy and complexity. This indicates the feasibility of applying the regression methods at constructing the statistical models of minerals processing.http://izvestiya.tpu.ru/archive/article/view/1734regression modelsimulation experimentselectionmulti-row polynomial algorithmmineralsore grinding |
spellingShingle | Seyran Shamirovich Balasanyan Hermine Mikhaylovna Gevorgyan Comparative analysis of the regression methods and data group accounting method in modeling mineral processing Известия Томского политехнического университета: Инжиниринг георесурсов regression model simulation experiment selection multi-row polynomial algorithm minerals ore grinding |
title | Comparative analysis of the regression methods and data group accounting method in modeling mineral processing |
title_full | Comparative analysis of the regression methods and data group accounting method in modeling mineral processing |
title_fullStr | Comparative analysis of the regression methods and data group accounting method in modeling mineral processing |
title_full_unstemmed | Comparative analysis of the regression methods and data group accounting method in modeling mineral processing |
title_short | Comparative analysis of the regression methods and data group accounting method in modeling mineral processing |
title_sort | comparative analysis of the regression methods and data group accounting method in modeling mineral processing |
topic | regression model simulation experiment selection multi-row polynomial algorithm minerals ore grinding |
url | http://izvestiya.tpu.ru/archive/article/view/1734 |
work_keys_str_mv | AT seyranshamirovichbalasanyan comparativeanalysisoftheregressionmethodsanddatagroupaccountingmethodinmodelingmineralprocessing AT herminemikhaylovnagevorgyan comparativeanalysisoftheregressionmethodsanddatagroupaccountingmethodinmodelingmineralprocessing |