Multi-attributive border approximation area comparison model based on Sugeno-Weber power fuzzy linguistic variables with Z-number information

Abstract

The tool of fuzzy Z-number linguistic term (FZNLT) set is significant and dominant because it integrates fuzzy set theory, linguistic set theory, Z-number theory, and uncertainty, allowing for massive accurate, and precise interpretation of incomplete, inconsistent, or vague data. It improves decision-making in problematic scenarios where traditional techniques may fall short. FZNLT set theory provides a valuable and better framework for modeling genuine-life problems with inherent ambiguity. Inspired by the above characteristics of the FZNLT set, our key goal is to design a constructive model of FZNLT sets, which is a modified form of many existing techniques. Also, construct the model of Sugeno-Weber operational laws based on valuable and dominant norms, called Sugeno-Weber t-norm (SWTN) and Sugeno-Weber t-conorm (SWTCN). Based on these operations, we analyze the power weighted aggregation operators for FZNLT information, with their certain basic properties. Additionally, the FZNLT-multi-attributive border approximation area comparison (FZNLT-MABAC) network based on Hamming distance measures for the above-proposed operators is developed. Ensemble learning is a dominant and critical model in machine learning technique that combines numerous techniques or models to enhance the performance or assessment of predictions, which is specially employed in complicated and problematic tasks such as early diabetes detection. For this, we demonstrate a numerical example to evaluate the application of ensemble learning approaches for early diabetes detection based on the above operators and construct the tool for the interpretation or comparison of the proposed theory and existing theory based on their ranking values to summarize the concluding remarks about proposed approaches.