An Investigation of the environmental sustainability index in terms of its prediction and clustering capabilities

During the years 1999-2005, the environmental sustainability index (ESI) constituted a predominant tool for evaluating, ranking, and grouping countries in terms of their current and future potential to protect the environment. In this piece of research, an investigation of the calculation/prediction, ranking, and clustering capabilities of the ESI 2005 is performed using traditional as well as computational intelligence tools, the latter including supervised general regression artificial neural networks, probabilistic artificial neural networks, unsupervised self-organizing maps, and fuzzy clustering. The results of the investigation shed some light on the derivation of the ESI, but further research is required for elucidating - and, thus, being able to replicate - the ESI values and clusters.