Abstract

GONZALES ARGOTE, Heynz Roberth  and  TICONA GONZALES, Ulises Amaru. Clustering, Landlockedness and International Trade: Empirical Application of the Partitioning Around Medoids and K-means algorithms. rlde [online]. 2019, n.32, pp.95-129. ISSN 2074-4706.

Landlockedness has generated significative interest in the geopolitical debate, particularly in Bolivia. This fact, along with innovative methodologies such as artificial intelligence and data mining, has motivated this research, which is unprecedented in the literature concerning landlockedness analysis through unsupervised algorithms of data mining. Consequently, the theory of cluster formation is studied and applied through the K-means and PAM (Partitioning Around Medoids) algorithms using international trade information of one hundred eighty-eight countries over a period of ten years, in order to test whether the landlockedness condition is a limiting factor in the commercial dynamics of countries. The results show that a reduced subset of the landlocked countries, including Bolivia, would have eased restrictions such as international trade costs and times, during the last decade.

Keywords : Cluster; landlocked countries; littoral; international trade; data mining.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )