Clustering analysis for Small Enterprises in Ecuador

Abstract

The proposed research aims and check the possibility of a cluster analysis to group small business according to the data defined as it appears in the Directory of Companies and Establishments 2013 Ecuador.
Using SPSS a total of 45 174 into the category of small companies were filtered. RapidMiner was employed to perform the clustering and k-means algorithm was used, that allowed to group into five clusters the total number of companies on the basis that the following indicators: 2103 net exports, sales 2013 2103 domestic sales, employees 2013.
A group of 159 companies with an average of 28.1 employees and annual exports in the order of around 616 371.51 USD were distinguished. . The study suggests conducting research, with companies selected according to the grouping obtained to try to categorize models from indicators that reflect the existing databases in Ecuador for different types of companies, to deepen in the factors that determine the small business success.