Predictive Modeling of Health Campaigns for Private Sector Companies in Perú Using the Automated Learning and OSEMN Method

Shimomura A.; Calderon S.; Wong L. · Applications in Software Engineering - Proceedings of the 13th International Conference on Software Process Improvement, CIMPS 2024 (2024) · DOI: 10.1109/cimps65195.2024.11095919

Resumen

Due to the new normal that the pandemic left us with, private companies have had to reinforce and improve the health campaigns implemented for the well-being of their collaborators. However, there is not enough justification as to which of these health campaigns are appropriate to implement in response to the need within companies. Therefore, a predictive model of health campaigns is proposed for a private company in Lima - Peru using 3 machine learning algorithms and the OSEMN methodology. The methodology consists of 5 phases: data acquisition, data cleaning, data exploration, algorithm modeling, and interpretation of results. The dataset used comes from 635 medical examinations of company employees, identifying 14 variables. The results showed that the proposed model obtained an AUC of 0.998, exceeding the optimal minimum percentage of 0.90.