Welfare analysis is a very important issue in the context of socio-economic. There is a huge literature about economic welfare and exist many conventional techniques and traditional methods that are used in almost all economic welfare studies. Regardless of the method used in welfare studies, the key element to promote the welfare of households in a community is to identify factors affecting welfare. For the first time, Adaptive Network Based Fuzzy Inference System (ANFIS) which is a powerful artificial intelligence tool is used as a new approach in economic welfare. To do this, real micro-data including some characteristics of households and housing units were used. We identified the most important factors affecting household welfare applying the ANFIS method. The ANFIS was then parameterized using these factors in order to predict the welfare measure. The empirical results showed that the ANFIS method outperforms the multiple regression method.
