Prototipo de Predicción del Desempleño Laboral Aplicando Técnicas de Minería de Datos

Abstract

The permanence of an employee in the company depends on many factors, the work climate, his/her personal goals, how comfortable he/she feels with the functions he/she performs, projections of promotions within the organization, among others. However, analyzing and controlling all these factors to ensure retention of human talent and ensuring that their capabilities are tapped into the right jobs is not an easy task. Every organization, however small, has relevant information about each one of its collaborators, as well as those aspiring to the vacancies of the company that at some point have gone through the selection process; data such as their personal information, family, work and academic history, as well as information about their skills, abilities and habits. But small businesses in general are unaware of the potential that this information can hold. Data mining, through its classification and prediction techniques, can be used to analyze information, extract knowledge and prediction models of future data trends, among the most used techniques are: decision trees, neural networks, Bayesian methods, genetic algorithms and fuzzy logic. This work intends to use the decision trees technique to design a model that uses the data that the company SFC Pack S.A.S. has stored in its human talent management processes, and with the help of the mentioned techniques, allows to anticipate the work performance of an aspirant to the position of higher staff turnover within the company, his/her adherence and permanence in the company and his/her probability of success in the execution of certain functions.