An environmental performance evaluation system: a neuro-fuzzy modeling approach

This paper deals with the definition of an environmental performance evaluation system of touristy and receptive facilities in rural Mediterranean areas. Such a system should act as a tool for design assistance, i.e. it could be useful to support building design choices, according to environmental sustainability. To derive the model of the evaluation system, a neuro-fuzzy approach is employed, that enables automatic extraction of the model from a large data set containing descriptors of the touristy buildings and related evaluation indexes. Specifically, the evaluation model is extracted through neural learning from data in form of fuzzy rules that can be easily interpreted and used by designers and architects to better understand the relationships existing between building features and environmental performances.