A practical approach for estimating weights of interacting criteria from profile sets

摘要: Aggregating multi-criteria data is an important problem with many applications. Commonly-used additive aggregation methods, such as the weighted arithmetic mean, cannot account for criteria interactions encountered in practical decision problems. The Choquet integral a suitable operator presence of interacting criteria. It replaces weight vector fuzzy measure that models importance each subset or coalition criteria, rather than just individual However, estimating measures practice has been problematic. Conventional approaches are cognitively challenging makers, while more recent suffer from prohibitive requirements. In this paper, we present formulation weights uses principal component analysis to interaction. This novel unsupervised approach required overcomes limitations other methods. applied two case studies environmental and sustainability analysis, results compared those mean. first triple bottom line 135 Australian industry sectors evaluated against 11 second life cycle assessment 8 alternative biosolids management options 5 These examples demonstrate ability proposed interaction these contexts.