Improving Military Recruit Quality Through Smart Classification Technology

摘要

The purpose of this research effort was to demonstrate the benefits of using sophisticated batch classification methods to improve the quality of enlisted Military personnel. In order to reach the objective, this project used the Belgian Psychometric Model as a technology demonstrator. The aim was to use this system with actual selection data from different countries to show that it can improve the quality of the enlisted persons through smart classification. In this report we review three datasets originating from respectively Portugal, Spain and the United States. The table below illustrates the core of what is meant by the title of this research project. The two compared classification methods deal with the same applicants for the same vacancies. Both methods fill the vacancies. Yet, the available indicators of the Military recruit quality (in this example, the relevant composite scores) are significantly better when the classification is done by a snrart classification system such as the Psychometric Model.