Оцінка впливу невизначеності на виникнення ризиків в ланцюгах постачання
Loading...
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Класичний приватний університет, Запоріжжя
Abstract
UKR: Метою досліджень є визначення та апробація підходу урахування впливу невизначеності на виникнення ризиків, з якими зіштовхуються при плануванні роботи ланцюгів постачання. При здійсненні досліджень в умовах виробничого підприємства визначено задачу прогнозування замовлень спеціальних профілів прокату для автомобільних доріг. Поєднання баєсовського підходу для детермінації задач невизначеності з ітераційним алгоритмом імпутації відсутніх даних дало можливість сформулювати рекомендації для вирішення подібних задач в практичних умовах. На підставі рівняння множинної регресії створено математичну модель, яка визначає потребу в продукції підприємства. Визначено прогнозні замовлення продукції підприємства. Залучення запропонованої моделі забезпечує скорочення економічних ризиків при роботі підприємства як ланки ланцюга постачання. Статистична оцінка результатів свідчить про їх надійність.
ENG: Russian aggression has destroyed much of Ukraine's infrastructure. The mechanism of functioning of supply chains has become unbalanced. Rapid response to challenges and identify cation of measures to prevent them is identified as a component of the Ukrainian economy in the struggle for victory. Uncertainty is understood as a situation when the probability of a certain event is not fully known. The import of missing data is considered to be a main approach to deal with the issue. The existence of uncertainty means the impossibility of clearly defining the future outcome and the need to respond to a threat that is likely to be an economic risk. The application of Monte Carlo methods in economics is not always acceptable to measure uncertainty. Classical probability theory has certain limitations for practical application to treat uncertainty. The Bayesian theory is applied as a preferable toolkit to address the issue. In practical terms, the main difficulty is to reduce the initial uncertainty given the results of statistical observations. The main approach to solving problems of uncertainty – imputation – enables to substitute the missing data with values determined by a certain algorithm. The main types of imputation are examined in the article. The goal of the research is to set forth an approach dealing with the impact of uncertainties on the economic risk of a supply chain link. The research was performed on the basis of a manufacturing enterprise. The task was to forecast orders for the supply of special metal shapes to fix the roads destroyed during the war. Due to the need to make decisions when potential consumers of its products have not yet decided on their orders for the supply of its products, the company faces uncertainty challenges. To solve this problem, the Expectation-Maximization algorithm was applied to treat to predict initially missed data. By obtaining full data due to the EM-algorithm under the uncertainty, an equation was obtained to predict the relationship between the technological parameters of roads and the future demand for the company`s products. The equation obtained enables to determine the forecast output of special profiles required to meet the need in rolled shapes. The statistical evaluation of research results testifies to their reliability.
ENG: Russian aggression has destroyed much of Ukraine's infrastructure. The mechanism of functioning of supply chains has become unbalanced. Rapid response to challenges and identify cation of measures to prevent them is identified as a component of the Ukrainian economy in the struggle for victory. Uncertainty is understood as a situation when the probability of a certain event is not fully known. The import of missing data is considered to be a main approach to deal with the issue. The existence of uncertainty means the impossibility of clearly defining the future outcome and the need to respond to a threat that is likely to be an economic risk. The application of Monte Carlo methods in economics is not always acceptable to measure uncertainty. Classical probability theory has certain limitations for practical application to treat uncertainty. The Bayesian theory is applied as a preferable toolkit to address the issue. In practical terms, the main difficulty is to reduce the initial uncertainty given the results of statistical observations. The main approach to solving problems of uncertainty – imputation – enables to substitute the missing data with values determined by a certain algorithm. The main types of imputation are examined in the article. The goal of the research is to set forth an approach dealing with the impact of uncertainties on the economic risk of a supply chain link. The research was performed on the basis of a manufacturing enterprise. The task was to forecast orders for the supply of special metal shapes to fix the roads destroyed during the war. Due to the need to make decisions when potential consumers of its products have not yet decided on their orders for the supply of its products, the company faces uncertainty challenges. To solve this problem, the Expectation-Maximization algorithm was applied to treat to predict initially missed data. By obtaining full data due to the EM-algorithm under the uncertainty, an equation was obtained to predict the relationship between the technological parameters of roads and the future demand for the company`s products. The equation obtained enables to determine the forecast output of special profiles required to meet the need in rolled shapes. The statistical evaluation of research results testifies to their reliability.
Description
І. Романовський: ORCID 0000-0002-2479-9457
Keywords
невизначеність, ризик, ланцюг постачання, втрати, сценарій, ланка, uncertainty, supply chain, imputation, missing data, algorithm, КЕП
Citation
Романовський І. Г. Оцінка впливу невизначеності на виникнення ризиків в ланцюгах постачання. Приазовський економічний вісник. 2022. № 2 (31). С. 74–78. DOI: https://doi.org/10.32840/2522-4263/2022-2-12.