PROCEDIMIENTO PARA EL ANÁLISIS DE LA ACCIDENTALIDAD LABORAL CON ÉNFASIS EN MODELOS MATEMÁTICOS/A PROCEDURE FOR THE ANALYSIS OF THE ACCIDENT RATE IN THE WORKPLACE WITH EMPHASIS ON MATHEMATICAL METHODS
2015
RESUMEN La investigacion fue aplicada en una empresa de mantenimiento y construccion de redes de generacion de energia electrica. El objetivo consistio en identificar el modelo de regresion que mejor ajusta a factores incidentes en la ocurrencia de accidentes laborales. Entre las tecnicas y herramientas que fueron utilizadas se encuentran las propias de la Gestion del Riesgo Laboral y las relacionadas con la estadistica descriptiva y multivariada. Al termino de la investigacion, se obtuvo que el analisis de diferentes modelos matematicos para explicar la accidentalidad laboral, permitio identificar que el de Regresion Logistica es aquel que brinda resultados con un menor margen de error. Se finalizo la investigacion proponiendose para estas variables, un programa de acciones de mejoras e indicadores que posibilitan el control de su desempeno. ABSTRACT The research was applied in an enterprise for the maintenance and construction of electricity generation networks. It is aimed at identifying the model of regression that better fits the incidental factors in the occurrence of accidents in the workplace. Among the techniques and tools used for the development of the research were found specific techniques related to the Labor Risks Management, and descriptive and multivariate statistics are applied. At the end of this research work, an examination of the different mathematical models to explain the accident rate in the workplace is obtained as the main result. This allows the author to identify that the Logistic Regression model is the one that provides the least margin of error. The research is ended by proposing a program of actions for the improvement of the accident rate in the workplace, as well as indicators that allow the control of its development.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
0
Citations
NaN
KQI