Work

Neuromorphic AI Chip Selection Using a Bipolar Intuitionistic Fuzzy BWM-MAIRCA Framework

Year: 2026
Type: article
Abstract: To address the increasing diversity of neuromorphic artificial intelligence (AI) chip architectures, whose technological capabilities and performance characteristics vary considerably across competing alternatives, this study develops an integrated bipolar intuitionistic fuzzy (BIF) Best-Worst Method (BWM) and Multi-Attributive Ideal-Real Comparative Analysis (MAIRCA) framework, termed BIF-BWM-MAIRCA, for neuromorphic AI chip selection in a multi-criteria group decision-making (MCGDM) environment under uncertainty. The framework uses bipolar intuitionistic fuzzy numbers to capture both positive and negative expert evaluations and a BIF weighted averaging operator to aggregate individual assessments into a collective decision matrix. BWM then derives the criterion weights, and MAIRCA ranks the competing alternatives based on the gaps between their theoretical and real performance. The framework's applicability is demonstrated through a case study in which three experts evaluate eight neuromorphic AI chips against five criteria. The results show the framework's ability to integrate uncertain expert evaluations, quantify the relative importance of the criteria, and prioritize alternatives within a single decision-making structure, while sensitivity and comparative analyses confirm the stability of the top-ranked alternative and the methodological advantages of the proposed approach. The framework thus provides a formal mathematical basis for selecting neuromorphic AI chips under bipolar uncertainty and conflicting expert preferences.
Source: The Nucleus
Language: English
Open Access status: diamond
OpenAlex ID: W7235182644
FWCI:
Cites: 38
Cited by:
Related to: 10