Merging Entropy in Self-Organisation: A Geographical Approach
Spatially-referenced data has achieved a unique place in regional science over the last decades. Much of the evolution Geographic Information Systems and Science have witnessed is due to the advances in the field of geocomputation and categorization of social and economic phenomena over geographical space. One of the traditional ways of analyzing socioeconomic data is by using rigid administrative boundaries, where internal structure, as well as the distribution of phenomena, lead to the disruption of their internal structure. This chapter assesses a more natural approach for data aggregation by using self-organizing maps. It aims to extend the debate on mutual information as well as spatial data, showing how data aggregation directly affects entropy values within the correlation of regions. This supports the identification of a new method that registers stronger correlated areas through a combination of entropy and self-organization, which offers new insights into topological innovation of spatially-explicit data and its integration in the field of regional science.
- Cover, T. M., & Thomas, J. A. (2006). Elements of information theory (2nd ed.). Hoboken, NJ: Wiley.Google Scholar
- Hausser, J., & Strimmer, K. (2009). Entropy inference and the James-stein estimator, with application to nonlinear gene association networks. The Journal of Machine Learning Research, 10, 1469–1484.Google Scholar