Creation of business profiles of exporting companies through unsupervised learning

This research presents a method for the creation of business profiles of export companies in the city of Barranquilla - Colombia. The method is supported by the development of unsupervised data learning techniques, Principal Component Analysis and Cluster analysis. The database used corresponds to primary financial information of 107 companies registered in the Chamber of Commerce of Barranquilla. First, a review of the context of global and local exports is presented. Next, the analysis of components allowed the creation of new uncorrelated variables and the joint visualization of variables in a two-dimensional space. Subsequently, the cluster analysis of the companies was interpreted in the cluster analysis according to the financial items studied. The work shows that the methodology employed allows identifying the type of service that an enterprise needs to achieve the best levels of performance, which in turns facilitates the implementation of development plans and of public and private policies of business promotion. © 2019 Centro de Informacion Tecnologica. All rights reserved.

• Autores: de la Hoz-Dominguez, Enrique J.; Escorcia-Guzman, Adalberto; Fontalvo-Herrera, Tomás J
• Fecha de publicación: 2019
• Institución: Universidad Tecnológica de Bolívar
• Repositorio: Repositorio Institucional UTB
• Idioma: spa
• OAI Identifier: oai:repositorio.utb.edu.co:20.500.12585/12336
• Acceso en línea: https://hdl.handle.net/20.500.12585/12336
• Palabra clave: Bankruptcy Prediction; Credit Scoring; Prediction; LEMB
• Rights: openAccess
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/