Fuzzy Clustering for Economic Data Mining: Mathematical Algorithmic Interpretations

Abstract

Fuzzy clustering techniques represent a natural extension of traditional clustering methodologies that have been developed to more accurately model uncertainty and imprecision in economic datasets. We argue that fuzzy clustering is well-suited to tackle many real-world economic problems, including, but not limited to, market segmentation and economic forecasting. Instead of forcing a data point would belong to only one cluster, the fuzzy clustering method allows a data point to belong to multiple clusters with a certain degree of membership, making them a more flexible technique compared to well-known hard clustering methods. Subsequently, the article discusses the problems and limitations of fuzzy clustering in economics, computational complexity, fuzzy parameters and uncertainty, and result interpretation.Contributions include fuzzy clustering to other machine learning methods, apply fuzzy clustering to big data, and fuzzy clustering to improve economy policy making. In conclusion fuzzy clustering represents a precious resource for in-depth and timely insights which can contribute to policy maker, business and researcher developments by discovering new target areas to invest, consequently increasing fountains of knowledge.