A conventional LASSO (least absolute shrinkage and selection operator) regression model utilizing the Pythagorean fuzzy sets in a system reliability analysis is developed. Overall, the Pythagorean fuzzy multivariate regression analysis enables decision makers to correctly identify the relationships between a set of responses in the form of fuzzy or non-fuzzy interpretive variables. The interpretability of the model is significantly improved by the proposed Pythagorean fuzzy LASSO regression model (PFLRM). Thus, a system reliability analysis is considered as an application of the study to evaluate the efficiency and effectiveness of the proposed PFLRM. There is no doubt that a system reliability analysis is vital to improve the safety performance of chemical processing industries, where an extensive number of industrial accidents occur annually. These accidents have subsequently highlighted the failure of some of the intervention actions to keep the systems safely in operation. The results illustrate a better performance with higher accuracy with the proposed PFLRM compared with the existing number of fuzzy regression models, particularly in the availability of non-informative variables.
|Number of pages||16|
|Journal||Neural Computing and Applications|
|Publication status||Published - Jul 2021|
- Pythagorean fuzzy LASSO
- Process industries
- Human reliability
- Corrective actions