Proposição de um modelo baseado em inferência neuro-fuzzy para segmentação de fornecedores sustentáveis
Due to the globalization of supply chains and the consequent increase in the quantity and diversity of suppliers, their segmentation has become fundamental, because it helps purchasing companies in the definition of specific strategies for suppliers that have similar characteristics. Given the need...
| Autor principal: | Saugo, Ricardo Antonio |
|---|---|
| Formato: | Dissertação |
| Idioma: | Português |
| Publicado em: |
Universidade Tecnológica Federal do Paraná
2022
|
| Assuntos: | |
| Acesso em linha: |
http://repositorio.utfpr.edu.br/jspui/handle/1/28496 |
| Tags: |
Adicionar Tag
Sem tags, seja o primeiro a adicionar uma tag!
|
| Resumo: |
Due to the globalization of supply chains and the consequent increase in the quantity and diversity of suppliers, their segmentation has become fundamental, because it helps purchasing companies in the definition of specific strategies for suppliers that have similar characteristics. Given the need to incorporate the concept of sustainability into supply chain management, economic, environmental and social performance criteria are also considered in the supplier assessment process. However, in the literature there are few works that present models for segmenting sustainable suppliers, and none of the published works uses supervised learning techniques. Therefore, the objective of this study is to propose a decision model for segmenting sustainable suppliers based on neuro-fuzzy inference systems (ANFIS). The proposed approach combines three ANFIS models in a three-dimensional quadratic matrix, based on several criteria associated with the dimensions of the triple bottom line. 108 candidate topologies were implemented with the help of the Neuro-Fuzzy Designer tool of the MATLAB® software. For the training and testing of these topologies, simulated samples from 200 supplier evaluations were used, generated with the help of the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method. The mean square error (MSE) between the desired values and the estimated values by each ANFIS model was calculated in order to select the best topologies and verify the accuracy of the models. The results provided by the topologies with the lowest mean squared error were analyzed using statistical tests. This study can be useful to help researchers and developers of computational solutions, mainly by providing adequate topological parameters to obtain accurate results in the application in question. |
|---|