kernel-based principal component analysis

This paper aimed to create a composite index for environmental quality. Since the results of empirical studies on economic - environmental variables nexus, considering different environmental indicators, are not consistent with each other; it seems necessary to use a comprehensive index that includes all dimensions of environmental pollution.

Using 6 environmental indicators related to two groups of selected OPEC and OECD countries for the period 2010 to 2019 and using three methods including principal component analysis, kernel-based principal component analysis, fuzzy robust principal component analysis, the creation of a composite environmental quality index is examined.

The results showed that FRPCA method has more efficiency and ability in weighting environmental indicators than other methods due to having the lowest error criteria. Therefore, using this method, the composite index was calculated. Moreover, the results showed that along with the upward trend of economic growth, the quality of the environment follows a downward trend in OPEC countries and an upward trend in OECD countries.

 Based on the results, policy recommendations as well as new perspectives and suggestions for future studies are presented.